ComfyUI Workflows
Browse real AI workflows from Flux, SDXL, Pony, ControlNet, video generation and commercial AI creators.
LTX 2.3 Workflows (RuneXX workflows, remade by Stefan Falkok) + NSFW (Base I2V, First-Last frame, Controlnet, Edit, Add Audio + Lipsync, Foley, Extended Video)
all base workflows for ltx 2.3, remade by me, creator is RuneXXMy ComfyUI Build - https://huggingface.co/datasets/StefanFalkok/ComfyUI_portable_torch_2.11.0_cu130_cp313_sageattention_triton/tree/mainMy NSFW TG Channel - https://t.me/stefanfalkokmothMy TG Channel - https://t.me/StefanFalkokAIMy TG Chat - https://t.me/+y4R5JybDZcFjMjFi
LTX 2.3 Seed Hunter Multiroll Workflow
Change the way you approach LTX genning. We all know that prompt adherence with LTX is notoriously bad. The model treats your prompts more like suggestions that it can ignore at will. So I say, stop beating your head on the wall trying to prompt better. Instead, the path to success with the model is to see the outcomes of many seeds quickly, and be able to choose one to bring to finalization. This is the main workflow I use to produce my short films, and I hope you all enjoy it too.See my full tutorial video showing how to use the workflow here:
LTX-2.3, simple workflow. T2V, I2V, I2V external audio, FFLF
Pretty simple workflow for LTXV with 3 modes so far and quite an early version. There is no prompt enchancer; in my opinion, it's useless garbage as a tool, and text LLM work mostly works very slow in ComfyUI. There is no latent upscaling. Almost the same useless garbage.Setting the full resolution will be faster and better than half resolution and latent upscaling with an interpolated result, but it requires resources like second generation. If your system is not powerful enough for the full resolution you set, almost certainly the server will run out of memory on the latent upscale stage. And that's even worse in IMG2video cases with the pixel upscaler itself.The workflow works fine on my 8GB VRAM videocard at 720p; I haven't tried higher resolution yet on FP8 dev model with distilled lora. There is a reason why VAE are separated into loaders and files, I haven't finished that part fully. Controlling it is the same as any other my workflow. If you try even one of my workflows, it will be similar.
![[LTX 2.3] 4-Image Reference to Video (Manual Control + OmniNFT LoRA)](https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/2911664d-7470-4be9-b81a-6a4bc744f189/original=true/132770096.jpeg)
[LTX 2.3] 4-Image Reference to Video (Manual Control + OmniNFT LoRA)
🚀 Run this Workflow for FREE on Runninghub: https://www.runninghub.ai/post/2061355424439431169/?inviteCode=rh-v1497 (🎁 Tip: Sign up via the link above to get 1,000 free credits + 100 daily login points!) 🎥 Watch the Full Video Tutorial & Review: 🌟 OverviewReady to push LTX 2.3 to its absolute breaking point? This workflow is designed to handle 4 reference images, attempting to mimic the advanced multi-image capabilities of closed-source giants like Grok.Because pushing 4 images significantly increases the risk of consistency breakdown, this workflow uses Manual Nodes for the reference inputs. This gives you absolute granular control over how the global prompts and segmented prompts interact.🚀 Key FeaturesManual prompt relay Node Architecture: Built for power users who want strict control over 4 separate image indexes and text prompts.OmniNFT LoRA Included: Essential for maintaining character body proportions when feeding the model too much visual data. It helps keep the aesthetics clean even under heavy load.
LTX 2.3 3-Reference Image-to-Video
🚀 Run this Workflow for FREE on Runninghub: https://www.runninghub.ai/post/2061372745287553025/?inviteCode=rh-v1497(🎁 Tip: Sign up via the link above to get 1,000 free credits + 100 daily login points!) 🎥 Watch the Full Video Tutorial & Review: 🌟 OverviewMissing the multi-image referencing features from closed-source models like Seedance and Grok? This experimental workflow pushes the open-source LTX 2.3 model to handle 3 reference images simultaneously.By utilizing the smart nodes from Prompt Relay, we assign equal weights (1.0) and sequential index frames to smoothly blend multiple visual concepts into a single video generation.🚀 Key FeaturesSmart Prompt Relay: Uses the intuitive Prompt Relay smart node for clean and easy 3-image setup.Open-Source Alternative: A great starting point to mimic commercial multi-image video models.
LTX 2.3 Image and Text Video 10S Similarity Preservation Workflow
Watch the full video first if you want to understand how this LTX 2.3 image-and-text video workflow works in practice. The video shows how one reference image can be combined with text control, how the 10-second similarity system keeps the subject stable, and how to run the full workflow online without rebuilding a complex local ComfyUI environment.This ComfyUI workflow is designed for LTX 2.3 image-reference video generation with text-controlled motion and 10-second likeness preservation. Its main purpose is to let creators start from one image, describe the desired action or camera movement with text, and generate a controlled video while keeping the original subject, composition, and visual identity more stable across the clip.The workflow is built around the LTX 2.3 distilled 1.1 generation route. It uses the LTX 2.3 video checkpoint, Gemma3 fp8 text encoder, LTX Audio VAE, LTXVConditioning, LTXVImgToVideoConditionOnly, LTXVPreprocess, Image_Resize_longsize, LTX2_NAG, ManualSigmas, CFGGuider, SamplerCustomAdvanced, LTXVLatentUpsampler, LTXVConcatAVLatent, LTXVSeparateAVLatent, tiled decoding, and final video output. This makes the workflow more structured than a basic one-pass image-to-video graph.The image side provides the visual anchor. The reference image is resized, prepared through LTXVPreprocess, and injected into the generation process through LTXVImgToVideoConditionOnly. This helps the model preserve the character, object, scene, lighting, clothing, and composition from the original image. The text prompt then controls the motion direction, expression, camera movement, atmosphere, and cinematic behavior.The key update is the 10-second similarity preservation system. The workflow uses similarity and anchor-style guidance during the later stages, especially around the latent upscaling and HD refinement process. This helps reduce common image-to-video issues such as face drift, hairstyle changes, clothing inconsistency, subject deformation, background collapse, and unwanted identity changes. For creators making character videos, this is one of the most important improvements.The generation process is divided into three stages. The first stage builds the initial composition and motion base. The second stage performs latent-space upscaling while keeping stronger similarity control and weak anchor stability. The third stage applies final high-definition refinement with lighter similarity control, improving sharpness and detail while trying not to damage the established character identity.The workflow also includes LTX2_NAG and a universal negative prompt system. This helps suppress flicker, frame jitter, subtitles, watermarks, UI overlays, bad hands, broken mouth shapes, unstable motion, unwanted text, distorted audio artifacts, and sudden scene changes. Compared with ordinary image-to-video workflows, this version is better suited for publishable creator content because it combines reference image control, text-guided direction, similarity locking, staged sampling, and high-resolution refinement.This workflow is suitable for character animation, portrait-to-video, product motion shots, cinematic still animation, AI short clips, MV fragments, social media video, Bilibili demonstrations, YouTube showcases, RunningHub releases, and Civitai workflow publishing.Main features:LTX 2.3 image-and-text video workflowOne reference image + text motion control10-second similarity preservationLTX 2.3 distilled 1.1 checkpoint routeGemma3 fp8 text encoderLTX Audio VAE supportImage_Resize_longsize image preparationLTXVPreprocess reference preprocessingLTXVImgToVideoConditionOnly image guidanceLTX2_NAG universal negative guidanceThree-stage rendering structureLTXVLatentUpsampler high-resolution transitionAV latent concatenation and separationFinal HD video outputSuggested workflow:Prepare one clean reference image first. The subject should be clear, well-framed, and not blocked by complex foreground objects. Load the image into the workflow, then write a text prompt describing the motion, camera behavior, lighting, expression, atmosphere, and video style. Run the first stage first to check whether the image identity and motion direction are correct. If the character changes too much, keep the 10S similarity settings active and simplify the prompt. If the video is too static, make the motion instruction more explicit. After the base motion is stable, continue through latent upscaling and final HD refinement.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/ai-detail/2061688885712875521?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1nVVr6QEd8/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/ai-detail/2061688885712875521?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1nVVr6QEd8/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 – Image/Text to Video – API Key Node Edition
Hello Community,Several users asked for my workflow, so I decided to share it with everyone. Maybe it will be useful for some of you as well.First Things First:The workflow is based on the original workflow available in ComfyUI → Templates → Video → Image to Video. I don't want to take credit for work that isn't mine. The original workflow structure comes from the ComfyUI team.What I Changed:I integrated the LTX API Node and removed the need for the heavy Gemma Text Encoder as well asthe Text Positive and Text Negative nodes. Since prompts can be entered directly into the API node, these additional nodes were no longer necessary. I also cleaned up the subgraph a bit to make it easier to understand and navigate (at least for me).Why I Made These Changes:The main reason was speed. By using the API node, the text encoding runs through the LTX server and is completed in just a few seconds. On my system this reduced the preparation time by roughly 10 minutes compared to my previous setup. For me, saving 10 minutes per generation is a significant improvement.Example Render Times:Measured on my system (see hardware section below) using 5-second clips:768 × 768 = 158.74 seconds1080 × 608 = 160.05 seconds1024 × 768 = 181.16 seconds1248 × 720 = 192.00 seconds1536 × 864 = 307.04 seconds1920 × 1080 = 459.06 secondsYour results may vary depending on hardware, settings, workflow configuration, and clip length.The workflow is also capable of generating 20-second clips at the above resolutions on my system.Hardware Used:AMD Ryzen 7 5700X64 GB DDR4 RAMNVIDIA RTX 4060 Ti 16 GBImage-to-Video and Text-to-Video:In the main view there is a switch node in the center of the workflow. Simply change it between True and False to switch between Image-to-Video (I2V) and Text-to-Video (T2V). No separate workflows are required.Important:Don't forget to enter your API key inside the subgraph. You can get a free API key here:https://docs.ltx.video/welcomeThen simply click "Get an API Key" (you have to sign up, but it's free).The required models are already listed on the left side of the workflow.Final Notes:This is not a completely new workflow. It is simply a practical modification of the original ComfyUItemplate that worked very well for me and significantly reduced my preparation time. Give it a try and let me know how it works on your system. Feedback, suggestions, and improvements are always welcome.
LTX-2.3 Long Video Generation (Distilled GGUF) workflows
AcknowledgementsI have addressed all known bugs within this workflow (excluding those inherent to ComfyUI or specific custom nodes). As such, I am concluding regular updates for this project.I might still push occasional updates if the mood strikes me, but for now, consider this the final version. Thank you so much for the downloads and support. I hope this workflow helps your creative process - happy generating!OverviewA simple LTX-2.3 workflow. It uses the Distilled GGUF model for fast generation.Core FeaturesSingle / Extend Generation ModeYou can choose whether to generate the video once or extend it afterward.Three prompt input modesPrompt enhancement using OllamaNative LTX prompt enhancementPlain (no enhancement)If Ollama is not needed, you can disconnect it from the Ollama SubGraph node or simply delete the node.Optional FeaturesPreview Switch: Displays a preview during sampling.Audio-driven: Generates a video that matches an existing audio file.T2V Switch: Ignores the start image and generates the video using text-to-video.Full Mode: Generated using Steps 15 and CFG 3.0. This takes a very long time. While the camera work and motion may be slightly improved, the generation time is not practical.Double-Frame Mode: For use with intense motion. By rendering at twice the frame rate, facial distortion is less likely to occur.NoteIf you encounter the error, please update ComfyUI-KJNodes to the latest version.
Friendly LTX-2.3 T2V+I2V+FLF+Lipsync
Welcome to my 💫🎦 Friendly LTX-2.3 T2V+I2V+FLF+Lipsync✨ Less mess, more magic· Unified VIBE FLF - Lipsync all-in one version including First Last Frame mode with latest quality enhancers and VibeVoice implementation for voice cloningUnified Omni - Lipsync all-in one version with latest quality enhancers and OmniVoice implementation for voice cloningUnified ID - Lipsync all-in one better quality version with two-stage video generation and ID-Lora implementation for voice cloningUnified Light - Lipsync all-in one version with single-stage video generationLTX-2.3 is a state-of-the-art DiT-based (Diffusion Transformer) audio-video foundation model developed by Lightricks. It represents a significant evolution from LTX-2, delivering enhanced audio and visual quality alongside improved prompt adherence. The model is specifically designed to generate synchronized video and audio within a single unified architecture, making it a powerful tool for multimodal content creation.I offer my unique workflow with convenient options control and all-in-one structure (audio, t2v, i2v)💻 System requirements:Minimum system requirements for 540p i2v and 720p t2v:RTX 3000-s, 8GB+ VRAM, 45GB+ RAM, 8-core processor, SSD, latest ComfyUI🚀 Low VRAM optional optimization:For systems with low VRAM use --reserve-vram ComfyUI parameter in run_nvidia_gpu.bat: --reserve-vram 4 (or other number in GB).📌 Detailed tips and links to models in the workflow✨ Workflow features:Extremely user-friendly interfaceMaximum performance and optimization from 8GB of VRAM: GGUF or 8-step distilled model with fp4 or fp8 text encoderAll-in-one: i2v, t2v, and interpolationConvenient one-click mode switchingPrompt enhancer and sampler previewsGeneration time setting in secondsLora support (up to 3)Detailed tips and links to all necessary modelsManual random seed for complete control over generations🤗🙏🏼 Thanks to Lightricks TeamOriginal repo — GitHub
LTX-2.3 DEV/DIST - IMAGE to Video and TEXT to Video with Ollama/RTX VSR/LTX-Director
V2.6 LTX-2.3 DEV & Distilled Video with Audio + LTX Director workflowMinor update to LTX2.3 Image to Video with Ollama workflow:sampling Preview implementedNAG to allow negative prompt for Distilled model (CFG=1)Included a Workflow supporting LTX Director node, which is awesome, it supports:First, Mid, Last or whatever frame, basically any frame is a keyframePrompt Relay which allows to generate frames prior to Input Imagemore precise editing: What shall happen when in the clip?Audio Import to create lipsynced clipsImage or Text to Video, even both in one processmore Info: https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUIrecommend to watch the creator´s youtube video in above link for how to.=> added a longer example video at the end.V2.5 LTX-2.3 DEV & Distilled Video with AudioImage to Video and a Text to Video workflow, both can use own Prompts or Ollama generated/enhanced prompts.works with latest LTX 2.3 Distilled model (8 steps, CFG=1) or Dev model (20 steps, CFG=3)Updated the processing for DISTILLED and DEV model, select the DIST or DEV model in loader node and switch to dedicated DIST or DEV processing pipeline, so each model has its own processing.DIST model pipeline: Standard Guider and Basic Scheduler, follows the manual sigmas issued by LightricksDEV model pipeline: MultiModal Guider and LTX Scheduler + Distilled Lora on latent upscalerIncluded a workflow version with "RTX Video Super Resolution" node, which upscales videos in highspeed.Tip: With latest Comfy and LTX updates, the processing got faster for me, so I can increase the scale_by in sampler node from 0.5 to 0.6 or higher to have crisper videos with minor impact on render time.V2.3 LTX-2.3 DEV & Distilled Video with AudioDownloads for LTX 2.3:update : April 14th 2026 : Lightricks has updated their LTX 2.3 distilled model to 1.1 (and Lora):Model (1.1 fp8 _scaled by Kijai): https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/diffusion_modelsdist. Lora 1.1 : https://huggingface.co/Lightricks/LTX-2.3/tree/mainLTX-2.3 Distilled & Dev Models (fp8_scaled): https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/diffusion_modelsTextencoder1: (fp8_e4m3fn, same as LTX-2): https://huggingface.co/GitMylo/LTX-2-comfy_gemma_fp8_e4m3fn/tree/mainTextencoder2: (projection_bf16): https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/text_encodersVideo & Audio Vae: https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/vaePreview VAE (taeltx2_3): https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/vaeLoras:Spartial upscaler (x2-1.1): https://huggingface.co/Lightricks/LTX-2.3/tree/mainDistilled Lora for upscaler (lora.384): https://huggingface.co/Lightricks/LTX-2.3/tree/mainSmaller, alternative Desitilled Lora by Kijai: https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/lorasDetailer Lora (same as LTX-2): https://huggingface.co/Lightricks/LTX-2-19b-IC-LoRA-Detailer/tree/mainOllama Model (prompt only, fast): https://ollama.com/mirage335/Llama-3-NeuralDaredevil-8B-abliterated-virtuosoalternative model with Vision (reads input image+prompt, slower): https://ollama.com/huihui_ai/qwen3-vl-abliteratedother model with Vision (great for I2V): https://ollama.com/huihui_ai/qwen3.5-abliteratedsmaller LTX 2.3 GGUF Dev or Dist. models work as well. (replace Checkpoint loader node with Unet loader node from this custom node: https://github.com/city96/ComfyUI-GGUF ):models: https://huggingface.co/unsloth/LTX-2.3-GGUF/tree/mainsave to models/unet/V1.5 LTX-2 DEV Video with Audio including latest 🅛🅣🅧 Multimodal GuiderImage to Video and a Text to Video workflow, both can use own Prompts or Ollama generated/enhanced prompts.Replaced the Guider node with latest Multimodal Guider node, see more details in WF notes or here: https://ltx.io/model/model-blog/ltx-2-better-control-for-real-workflows Before we had 1 CFG parameter for audio and video. With multimodal guider, we now can tweak audio and video seperately with even more parameters...added a Power Lora Loader node to inject further Lorasuse Image to Video Adapter Lora to improve motion for I2V: https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/tree/mainreplaced a node to no longer require comfymath custom nodesV1.0 LTX-2 DEV Video with Audio:Image to Video and a Text to Video workflow with own Prompts or Ollama generated/enhanced prompts.setup for the LTX2 Dev model.uses Detailer Lora for better quality and LTX tiled VAE to avoid OOM and visual grids2 pass rendering (motion+upscale). Upscale process uses distilled and spatial upscale Lorasetup with latest LTXVNormalizingSampler to increase video & audio quality.Text to Video can use dynamic prompts with wildcards.Download LTX-2 Files: (Workflow V1.0 and V1.5 only)Find Model/Lora Loader nodes within Sampler Subgraph node.- LTX2 Dev Model (dev_Fp8): https://huggingface.co/Lightricks/LTX-2/tree/main- Detailer Lora: https://huggingface.co/Lightricks/LTX-2-19b-IC-LoRA-Detailer/tree/main- Distilled (lora-384) & Spatial upscaler Lora: https://huggingface.co/Lightricks/LTX-2/tree/main- VAE (already included in above dev_FP8 model, but needed if you go for GGUF models): https://huggingface.co/Lightricks/LTX-2/tree/main/vae- Textencoder (fp8_e4m3fn): https://huggingface.co/GitMylo/LTX-2-comfy_gemma_fp8_e4m3fn/tree/main- Image to Video Adapter Lora (more motion with I2V): https://huggingface.co/MachineDelusions/LTX-2_Image2Video_Adapter_LoRa/tree/mainSave Location:📂 ComfyUI/├── 📂 models/│ ├── 📂 checkpoints/│ │ ├── ltx-2-19b-dev-fp8.safetensors│ ├── 📂 text_encoders/│ │ └── gemma_3_12B_it_fp8_e4m3fn.safetensors│ ├── 📂 loras/│ │ ├── ltx-2-19b-distilled-lora-384.safetensors│ └── 📂 latent_upscale_models/│ └── ltx-2-spatial-upscaler-x2-1.0.safetensors│ └── 📂 Clip/│ └── ltx-2.3_text_projection_bf16.safetensorsCustom Nodes used:https://github.com/Lightricks/ComfyUI-LTXVideohttps://github.com/rgthree/rgthree-comfyhttps://github.com/yolain/ComfyUI-Easy-Usehttps://github.com/stavsap/comfyui-ollamahttps://github.com/kijai/ComfyUI-KJNodeshttps://github.com/Comfy-Org/Nvidia_RTX_Nodes_ComfyUI (RTX VSR Version)Text 2 Video only: https://github.com/KoinnAI/ComfyUI-DynPromptSimplifiedLTX Director only: https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUIOllama help:Install Ollama from https://ollama.com/download a model: Go to a model page, chose a model , then hit the copy button, i.e. https://ollama.com/huihui_ai/qwen3-vl-abliteratedopen terminal and paste the model name, i.e.: ollama run huihui_ai/qwen3-vl-abliteratedmodel will be downloaded and can be selected in green comfy node "Ollama Connectivity". Hit "Reconnect" to refresh.Example longer Video
LTX 2.3 Director
Okay, This straight up is the coolest thing I've seen in a while.Thansk to What Dreams Cost for this insane set of nodesWith this node you can create T2V, I2V, add sound, combine, set keyframes, and basically driect your own video, all within one workflowrequirements:If you can run LTX2.3, you should be able to run thisAt least 12gb VRAMNodes and models are listed in the workflowWatch the video here:Instagram: https://www.instagram.com/synth.studio.models/Buy me a☕ https://ko-fi.com/lonecatoneThis represents many hours of work. If you enjoy it, please 👍like, 💬 comment , and feel free to ⚡tip 😉
PlagueKind Nodes - LTX Compatible LoRA Stack + Unified Resize Tools (ComfyUI)
note: this model will only be updated when an entirely new node has been released. get incremental updates via comfy manager or git pull.ComfyUI-PlagueKind-NodesA utility node pack for ComfyUI focused on structured LoRA stacking and unified image/mask resizing for consistent workflow behavior across models.Included NodesLTX LoRA Loader StackA 10-slot LoRA stacking system designed for advanced workflows, including LTX models, while remaining fully compatible as a standard LoRA loader for other architectures.Features10 independent LoRA slotsEnable / disable per slotPer-slot strength controlOptional split control for LTX models:Video branch multiplierAudio branch multiplierWorks as a standard LoRA loader for non-LTX modelsSearchable LoRA selection menuFolder-aware organization and displayMissing file detection warningDrag and drop slot reorderingCompact UI designed for mobile and desktop useBehaviorFor non-LTX models, the node functions as a normal LoRA stacker using combined strength values.For LTX workflows, it supports separated audio and video branch weighting when applicable.Unified Resize Image / MaskA single consistent resizing node for both images and masks using a unified geometric pipeline.FeaturesMultiple scaling modes:Dimensions (W × H)MultiplierLonger SideShorter SideTotal Pixels (Megapixels)Aspect ratio preservation optionCenter crop modeDivisible-by constraint (useful for latent-based workflows like LTX / SDXL)Unified image + mask transformation pipelineStable tensor-based mask resizing (no mismatch between image and mask geometry)PurposePrevents common ComfyUI issues such as:image and mask misalignment after resizinginconsistent crop behavior between nodesunstable mask scaling in inpainting workflowsInstallationOption 1 (Recommended)Install via ComfyUI ManagerOption 2 Manual Installcd ComfyUI/custom_nodes git clone https://github.com/PlagueKind/Comfyui-PlagueKind-Nodes.git Restart ComfyUI.RequirementsNo external dependencies required beyond standard ComfyUI installation.Uses:torchcomfy.utilsNotesLoRA Loader Stack is fully compatible with standard LoRA workflowsLTX branch controls only apply when used with compatible modelsUnified Resize ensures identical transformations for images and masksLicenseMIT LicenseSupportIf you find this useful, consider supporting development:Monero (XMR):865BrcfWLdwELwuq5faV1uVTbh93zVK6AUYLY2c3mX6sFfAGRfS6axe1kBTYYKuM7ccN7zBZDAZvnT7E4NKmUazySdbpc7p
Video Outpainting - fast and simple with LTX-2.3
ComfyUI LTX-2.3 Video OutpaintingGithubThis is a lightweight ComfyUI workflow meant to quickly expand the viewport of a video and a minimum of fuss. Just load your video and select the new output area.Custom Node DependenciesI try to keep the external dependencies to a minimum, but there are a few required packages for this workflow to support the core functionality.ComfyUI-LTXVideoSearch LTXVideo in ComfyUI Manager.Provides the 🅛🅣🅧 Add Video IC-LoRA Guide node required to use the outpainting lora.comfy-mtbSearch mtb in ComfyUI Manager.Provides the Color Correct (mtb) node.Only required if you need the gamma correction feature of this workflow.Gamma correction can be used to temporarily brighten dark input video to help the Lora correctly identify which areas to paint. Note about this from the Lora developer.KJNodesSearch kjnodes in ComfyUI Manager.Provides get/set nodes and LTX2 Sampling Preview Override for sampler previews.Wan VACE Prepv1.0.19 or higher requiredSearch Wan VACE Prep in ComfyUI ManagerProvides the 🪐 Video Outpaint node.ComfyUI Nodes 2.0 : The Video Outpaint node has not been well tested under ComfyUI's Nodes 2.0 renderer and may or may not work correctly with it. No effort will be spent to ensure Nodes 2.0 compatibility or stability until Comfy publishes developer documentation for Nodes 2.0.ModelsIC-Lora Outpaint ModelDownload and place in your ComfyUI/loras directory.This lora makes the entire workflow possible.Changelogv1.0.0 Initial release.
LTX 2.3 Audio-Reactive Animation Workflow
Watch the full video first if you want to understand how this LTX 2.3 audio-reactive animation workflow works in practice. The video shows how one image and one audio track can be connected into a staged animation pipeline, how the video length follows the audio duration, and how to launch the workflow online without rebuilding the full ComfyUI environment locally.This ComfyUI workflow is designed for LTX 2.3 audio-reactive animation generation. Its main purpose is to take a source image and an audio file, then generate a video clip whose duration and visual rhythm are organized around the audio input. Instead of creating a silent image-to-video clip and manually matching it to music later, this workflow brings audio into the generation structure from the beginning, making it more suitable for music animation, MV fragments, sound-driven visual clips, and social media video production.The workflow is built around the LTX 2.3 video generation route. It uses image reference preparation, audio duration detection, automatic frame calculation, LTXVImgToVideoConditionOnly, LTXVConditioning, CFGGuider, ManualSigmas, SamplerCustomAdvanced, LTXVLatentUpsampler, AV latent combination, tiled VAE decoding, and CreateVideo output. The graph also includes VRAM purge tools, fps control, latent size checking, image resizing, mask handling, and multi-stage refinement.The audio side is one of the most important parts of this workflow. The input audio is measured through an Audio Duration node, then converted into a frame count through a math expression. This keeps the generated video length aligned with the audio length and reduces manual calculation errors. The workflow uses 24 fps logic and LTX-friendly temporal length rules, so the video can follow a cleaner generation structure instead of using arbitrary frame counts.The image side provides the visual identity. The source image is resized and prepared, then injected into the video process through LTXVImgToVideoConditionOnly. This allows the generated animation to preserve the original character, object, scene, or visual style while still producing motion. The same image reference can be reused across later refinement stages, helping the workflow maintain continuity after latent upscaling.The generation pipeline uses a three-stage structure. The first stage builds the initial animation and base composition. The second stage performs latent-space upscaling and refinement. The third stage applies final high-resolution polish before decoding and video output. This is more stable than a single-pass workflow because each stage has a clear purpose: establish motion, improve structure, then refine quality.Compared with ordinary image-to-video workflows, this graph is more useful for audio-based content. A normal I2V workflow may create motion, but the clip length and final output often need to be adjusted manually afterward. This workflow connects audio duration, frame calculation, image guidance, staged sampling, and final video output into one pipeline. It is especially useful for short MV visuals, music-reactive character clips, animated cover art, rhythm-based visual experiments, AI music videos, Bilibili demonstrations, YouTube content, RunningHub showcases, and Civitai workflow examples.Main features:LTX 2.3 audio-reactive animation workflowOne image + one audio inputAudio duration detectionAutomatic frame count calculation24 fps generation logicImage-guided video animationLTXVImgToVideoConditionOnly reference controlThree-stage rendering structureManualSigmas and SamplerCustomAdvanced controlLTXVLatentUpsampler high-resolution refinementMask and latent handling for staged consistencyCreateVideo final output with audioSuggested workflow:Prepare a clean source image and a clear audio file first. The image should have a readable subject, strong composition, and enough visual detail for animation. The audio should have stable volume and a clear rhythm or atmosphere. Load the image and audio into the workflow, then let the audio duration node calculate the target video length. Write a prompt that describes the motion, camera behavior, visual mood, and how the subject should respond to the music. Start with a short test to check whether the animation length, identity preservation, and motion direction are correct. If the motion feels too weak, make the prompt more explicit. If the image drifts too much, strengthen the reference structure and simplify the motion language. Once the first stage is stable, continue through latent upscaling and final refinement.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058938733910650882?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1gLGo6kEbF/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058938733910650882?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1gLGo6kEbF/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
Stable Audio 3 Sound Asset Generation Workflow
Watch the full video first if you want to understand how this Stable Audio 3 workflow works in practice. The video shows how a simple text idea can be expanded into a structured audio prompt, how different sound categories affect the result, and how to launch the workflow online without building the full ComfyUI audio environment locally.This ComfyUI workflow is designed for Stable Audio 3 sound asset generation. Its main purpose is to turn text descriptions into usable audio assets, including music tracks, instrument loops, sound effects, one-shot samples, ambience, cinematic hits, UI sounds, game audio elements, and production-ready creative sound material. Instead of only generating random audio from a short prompt, this workflow adds a category-aware prompt expansion layer so the final Stable Audio prompt becomes more precise and more suitable for the target audio type.The workflow is built around the Stable Audio 3 Medium Base route. It uses stable_audio_3_medium_base.safetensors as the main checkpoint, t5gemma_b_b_ul2.safetensors as the Stable Audio text encoder, and a separate text-generation route for intelligent prompt rewriting. The generation pipeline uses CLIPTextEncode for positive and negative conditioning, EmptyLatentAudio for defining the target audio duration, KSampler for latent audio sampling, and VAEDecodeAudio to decode the generated latent into an actual audio waveform.The most important design in this workflow is the optional reprompt system. Users can input a short idea, then decide whether to enable prompt expansion. When reprompt is enabled, the workflow uses a category-aware prompt template. The available categories include Music, Instrument, SFX, and One-shot. Each category has different prompt rules. Music prompts focus on genre, instruments, layers, rhythm, mood, BPM, and track length. Instrument prompts focus on playing technique, timbre, production texture, BPM, and loop or stem length. SFX prompts focus on sound source, material texture, spatial environment, movement, attack, decay, and duration. One-shot prompts focus on short isolated audio samples such as hits, stabs, plucks, drum sounds, impacts, or short sound design elements.This structure makes the workflow more practical than a simple text-to-audio setup. In ordinary audio generation, a vague prompt like “dark cinematic sound” may produce inconsistent results. Here, the same idea can be expanded into a more technical and production-oriented prompt, including instrument details, ambience, rhythm, physical texture, stereo space, and length. That gives the Stable Audio model clearer instructions and makes the output easier to use in real projects.This workflow is suitable for video creators, AI filmmakers, game developers, music producers, sound designers, short drama editors, YouTube creators, Bilibili creators, RunningHub users, and Civitai workflow collectors. It can be used to create background music, transition sounds, UI feedback, horror stingers, cinematic impacts, ambience beds, Foley-style effects, instrument loops, and short production samples.Compared with ordinary audio workflows, this version is more structured, easier to control, and better for repeated asset production. You do not need to manually write a professional audio prompt every time. You can start with a rough idea, choose the audio category, set the duration, control the seed, and let the workflow generate a more useful Stable Audio prompt before sampling.Main features:Stable Audio 3 sound asset generation workflowText-to-audio generation inside ComfyUIStable Audio 3 Medium Base checkpoint routeT5Gemma Stable Audio text encoderOptional intelligent reprompt systemMusic / Instrument / SFX / One-shot category presetsUser input replacement and audio length insertionEmptyLatentAudio duration controlKSampler latent audio generationVAEDecodeAudio waveform decodingSuitable for music, ambience, SFX, loops, and one-shot samplesOnline RunningHub execution without local setupSuggested workflow:Start with a short audio idea first. Choose the category that matches your target output: Music for full tracks, Instrument for loops or stems, SFX for environmental or action sounds, and One-shot for short isolated samples. Set the duration according to the asset type. Enable reprompt if you want the workflow to expand your rough idea into a more detailed technical audio prompt. If you already wrote a strong prompt yourself, disable reprompt and send the text directly into Stable Audio. Run a first seed test, listen carefully, then adjust category, duration, BPM language, instrument description, texture, or spatial details until the generated asset matches your production need.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058938930971635714?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1gLGo6kEbF/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058938930971635714?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1gLGo6kEbF/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 12GB GGUF DIRECTOR
One node to rule them all... basically!Here we have the usual 12GB GGUF workflow but, with a new Director node! This node allows you to fully control your generations. Like a video timeline editor you can load in images, place text prompts, add custom audio and so much more. The node is not made by me but by WhatDreamsCost, they have made a wonderful node that really simplifies generations and adds much more control to non power users and power users alike.You can find a tutorial video on the Github page, don't forget to give the repo a star to show your support for this fine dev!https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUIThis node is a WIP, the dev is adding video support and built in IC lora support and more.For now it replaces the t2v, i2v, ia2v, ta2v, and even multi keyframe workflows. So, you really only need this one and v2v for all the most basic of generation techniques in LTX. Reduces it all down to about 2 workflows and soon it will be all in one and not a huge workflow either.
LTX 2.3 Text to Video OmniNFT + Relay Three-Stage No-Subtitle Workflow
Watch the full video first if you want to understand how this LTX 2.3 text-to-video workflow works in practice. The video shows how a clean prompt can be turned into a complete video clip, how the three-stage rendering structure improves stability, and how to launch the workflow online without rebuilding the full ComfyUI environment locally.This ComfyUI workflow is designed for LTX 2.3 text-to-video generation, using OmniNFT, Relay-style prompt control, and the distilled 1.1 model route to create clean video outputs from text prompts. The main purpose of this workflow is to make text-to-video generation more controllable, more stable, and more suitable for publishing, especially when users want a no-subtitle, no-watermark, no-extra-text output.The workflow is built around the LTX 2.3 distilled 1.1 generation route. It uses an LTX 2.3 checkpoint, Gemma3-based text encoding, LTXVConditioning, EmptyLTXVLatentVideo, LTXVEmptyLatentAudio, LTX2_NAG negative guidance, ManualSigmas, CFGGuider, SamplerCustomAdvanced, LTXVLatentUpsampler, VAEDecodeTiled, LTXVAudioVAEDecode, and CreateVideo. The graph also includes seed control, fps control, universal negative prompting, VRAM management, and audio-video latent handling.The core idea is to generate video from text while maintaining stronger control over structure, motion, and final image quality. The positive prompt defines the subject, action, camera movement, lighting, environment, atmosphere, and cinematic direction. The negative prompt is designed to suppress common LTX video problems, including low quality, flicker, unstable perspective, identity drift, broken anatomy, subtitles, captions, UI overlays, logos, watermarks, unreadable text, and unwanted audio artifacts.The workflow uses a three-stage rendering structure. The first stage focuses on initial composition and motion foundation. It creates the base video latent and establishes the main visual direction. The second stage performs latent-space upscaling and refinement, allowing the workflow to improve structure and detail without rebuilding the whole video from scratch. The third stage applies final high-resolution polish, using another controlled sampling pass before tiled VAE decoding and video assembly.Compared with ordinary text-to-video workflows, this graph is more production-oriented. A simple one-pass T2V workflow may be fast, but it often suffers from weak motion, unstable composition, flicker, random text artifacts, or poor detail. This workflow separates generation into clear stages, uses negative guidance to suppress unwanted subtitles and watermarks, and applies latent upscaling before final output. That makes it more useful for creators who need cleaner video results for tutorials, showcases, social media, and workflow publishing.This workflow is suitable for cinematic shots, AI short clips, fantasy scenes, product-style motion, character motion tests, visual concept videos, MV fragments, Bilibili demonstrations, YouTube content, RunningHub showcases, and Civitai workflow examples. It is especially useful when you want to start from pure text but still keep the output clean and structured.Main features:LTX 2.3 text-to-video workflowOmniNFT + Relay-style prompt controlDistilled 1.1 model routeClean no-subtitle / no-watermark output directionGemma3 text encoder routeLTXVConditioning at controlled frame rateEmpty video latent and audio latent structureLTX2_NAG negative guidance supportUniversal negative prompt for clean video outputThree-stage rendering pipelineLTXVLatentUpsampler high-resolution refinementVAEDecodeTiled and CreateVideo final outputSuggested workflow:Start with a clear text prompt. Define the subject, main action, camera movement, lighting, environment, and desired video style. Keep the first test short and avoid overloading the prompt with too many competing actions. If the output contains unwanted text, subtitles, logos, or unstable artifacts, strengthen the negative prompt and simplify the scene. If the motion is too weak, make the action and camera direction more explicit. If the composition is good but the detail is not enough, continue through the latent upscaling and final refinement stages. Once the three-stage result is stable, export the video and use it directly for publishing or further editing.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2057671098963161090?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2057671098963161090?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Video Re-Speaking OmniNFT + Relay Lip-Sync Replacement Workflow
Watch the full video first if you want to understand how this LTX 2.3 video re-speaking workflow works in practice. The video shows how an existing talking video can be guided by a new audio track, how the lip-sync control pipeline is organized, and how to launch the workflow online without rebuilding the full ComfyUI environment locally.This ComfyUI workflow is designed for LTX 2.3 video re-speaking, word replacement, and audio-driven lip-sync generation. The main purpose of this workflow is to take an existing talking video or character reference and regenerate the mouth movement so it can match a new audio track. Instead of creating a completely new character from scratch, the workflow focuses on preserving the original person, framing, motion style, and video identity while changing the spoken content.The workflow is built around the LTX 2.3 distilled 1.1 route. It uses ltx-2.3-22b-dev-dare-ties-distilled-1.1 as the main checkpoint, Gemma3 fp8 text encoding, LTX Audio VAE, LipDub IC LoRA, VBVR / I2V stabilization LoRA, LTXVAudioVAEEncode, LTXVSetAudioRefTokens, LTXAddVideoICLoRAGuide, LTXVCropGuides, ManualSigmas, CFGGuider, SamplerCustomAdvanced, LTXVLatentUpsampler, LTXVAudioVAEDecode, and final video output.The key control module is LipDub IC LoRA. This module helps the model focus on mouth movement, speaking rhythm, and audio-related facial changes. The workflow encodes the new audio into audio latent space, then uses LTXVSetAudioRefTokens to inject audio reference tokens into the conditioning. This allows the rendering stages to follow the replacement speech instead of only producing generic motion.The video reference is handled through LTXAddVideoICLoRAGuide. This node injects the original video or visual reference into the generation process, helping the output preserve the face, camera angle, clothing, background, and overall identity. The workflow uses this guide across multiple rendering stages, so the character does not drift too far while the mouth movement is updated.The generation process is divided into three stages. The first stage creates the base lip-sync motion and audio-aligned structure. The second stage inherits audio tokens from the first stage and performs latent-space refinement. The third stage uses the previous output as a stronger reference and applies final high-resolution refinement. This staged structure is important because lip-sync generation is sensitive: a one-pass workflow can easily produce unstable mouths, drifting faces, flicker, or weak audio alignment.Compared with ordinary image-to-video or video-to-video workflows, this graph is more specialized for re-speaking. A normal V2V workflow may preserve motion but may not properly follow the new speech. A normal talking-head workflow may follow audio but may weaken the original video identity. This workflow combines video reference guidance, LipDub control, audio latent tokens, negative guidance, staged rendering, and latent upscaling to balance identity preservation and mouth synchronization.This workflow is suitable for dialogue replacement, AI dubbing previews, character re-speaking, translated video demonstrations, virtual host correction, short-form talking clips, product explanation videos, Bilibili demonstrations, YouTube content, RunningHub showcases, and Civitai workflow examples.Main features:LTX 2.3 video re-speaking workflowReplace speech while preserving video identityOmniNFT + Relay-style prompt controlDistilled 1.1 model routeLipDub IC LoRA for mouth and speech controlLTXVAudioVAEEncode for new audio encodingLTXVSetAudioRefTokens audio token injectionLTXAddVideoICLoRAGuide video reference guidanceLTXVCropGuides stage alignmentThree-stage rendering pipelineLTXVLatentUpsampler high-resolution refinementFinal audio decode and video outputSuggested workflow:Prepare a clean source video first. The face should be visible, the mouth area should not be blocked, and the camera should not shake too aggressively. Then prepare a clean replacement audio file with stable volume and clear speech. Load the video reference and the new audio into the workflow, then use a prompt that describes the speaker, lighting, framing, and natural speaking behavior. Run a short test first to check mouth alignment, identity preservation, and facial stability. If the mouth does not follow the audio strongly enough, increase the lip-sync guidance strength or simplify the visual prompt. If the face changes too much, reduce aggressive motion language and rely more on the original video guide. Once the first-stage result is stable, continue through the second and third refinement stages for cleaner output.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2057730271897800705?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2057730271897800705?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Single-Person Digital Human OmniNFT + Relay Audio-Driven Workflow
Watch the full video first if you want to understand how this LTX 2.3 single-person digital human workflow works in practice. The video shows how one character image and one audio file can be turned into an audio-driven talking video, how the staged rendering structure improves stability, and how to launch the workflow online without rebuilding the full ComfyUI environment locally.This ComfyUI workflow is designed for LTX 2.3 single-person digital human video generation, using OmniNFT, Relay-style prompt control, and the distilled 1.1 model route to create an audio-driven talking character video from a still image. The main purpose of this workflow is to make single-character digital human production easier, more repeatable, and more suitable for real creator use.The workflow starts with one character image. The image is resized and prepared before entering the LTX video pipeline. This image becomes the main identity reference for the digital human, controlling the face, clothing, framing, visual style, and overall composition. The workflow then uses LTXVImgToVideoConditionOnly to inject the image into the video latent process, allowing the model to preserve the original subject while generating motion.The audio side is also important. The workflow includes Audio Duration detection and a SimpleMath frame calculation system. The audio length is read automatically, then converted into an LTX-compatible frame count. This helps reduce manual frame-count mistakes and keeps the generated video length closer to the input audio. The workflow also uses LTXVEmptyLatentAudio, LTXVConcatAVLatent, and LTXVSeparateAVLatent to connect audio latent logic with video latent generation.The core generation process is divided into three stages. The first stage establishes the character motion, composition, and basic video structure. The second stage performs latent-space refinement and continuation. The third stage applies high-resolution refinement after latent upscaling. This three-stage structure is more stable than a simple one-pass render because each stage has a clearer purpose: build motion, improve structure, then polish quality.The workflow also includes LTX2_NAG and a universal negative prompt structure. These are used to reduce common digital human problems such as identity drift, face distortion, broken mouth shapes, unstable lip movement, flicker, frame jitter, unwanted subtitles, watermarks, bad hands, or sudden scene changes. For digital human videos, this is especially important because small facial errors are very noticeable.Compared with ordinary image-to-video workflows, this graph is more suitable for talking character production. A basic I2V workflow can create motion, but it may not properly handle audio duration, frame alignment, staged refinement, or identity preservation. This workflow combines image guidance, audio-aware frame logic, negative guidance, latent upscaling, tiled decoding, and final video assembly into one more practical pipeline.This workflow is suitable for AI presenters, single-person digital humans, virtual hosts, narration avatars, character dialogue clips, product explanation videos, short-form talking videos, Bilibili demonstrations, YouTube content, RunningHub showcases, and Civitai workflow examples.Main features:LTX 2.3 single-person digital human workflowOne image + one audio inputOmniNFT + Relay-style prompt controlDistilled 1.1 model routeAudio duration detectionAutomatic LTX-compatible frame calculationLTXVImgToVideoConditionOnly image guidanceLTXVEmptyLatentAudio audio latent structureLTXVConcatAVLatent and LTXVSeparateAVLatentLTX2_NAG negative guidance supportThree-stage rendering pipelineLTXVLatentUpsampler high-resolution refinementTiled VAE decoding and CreateVideo outputSuggested workflow:Prepare a clean single-person character image first. The face should be clear, the mouth area should not be blocked, and the subject should not be too small in the frame. Then prepare a clean audio file with stable volume and limited background noise. Load the image and audio into the workflow, then write a prompt describing the character, camera framing, lighting, expression, and speaking style. Start with a short test to check identity stability, mouth movement, and motion quality. If the face changes too much, reduce aggressive motion language and strengthen image guidance. If the result is too static, add subtle head movement or natural speaking motion to the prompt. Once the base motion is stable, continue through latent upscaling and final high-resolution output.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058190932108992514?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058190932108992514?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Image to Video OmniNFT + Relay One-Image Film Workflow
Watch the full video first if you want to understand how this LTX 2.3 image-to-video workflow works in practice. The video shows how one image can be turned into a complete video clip, how the staged rendering pipeline improves stability, and how to launch the workflow online without rebuilding the full ComfyUI environment locally.This ComfyUI workflow is designed for LTX 2.3 image-to-video generation, using OmniNFT, Relay-style prompt control, and the distilled 1.1 model route to turn a single image into a finished video clip. The main purpose of this workflow is to make one-image video generation more stable, more controllable, and more production-ready than a basic image-to-video graph.The workflow starts from a single input image. The image is resized and prepared through Image_Resize_longsize and LTXVPreprocess, then passed into LTXVImgToVideoConditionOnly as the main visual condition. This allows the source image to guide the video identity, composition, subject placement, and overall visual style while still giving the LTX model enough freedom to generate motion.The workflow is built around the LTX 2.3 distilled 1.1 route. It uses LTX video conditioning, LTX audio-video latent logic, an LTX Audio VAE, LTX2_NAG negative guidance, a universal negative prompt, and a three-stage rendering pipeline. The graph also includes Seed Everywhere, fps control, EmptyLTXVLatentVideo, LTXVEmptyLatentAudio, LTXVConcatAVLatent, LTXVSeparateAVLatent, ManualSigmas, CFGGuider, SamplerCustomAdvanced, LTXVLatentUpsampler, VAEDecodeTiled, and final video output.The key generation structure is divided into three stages. The first stage focuses on initial composition and base motion. It uses the image condition to establish the main character or scene and generate the first stable video latent. The second stage performs latent-space expansion, reconditioning, and continuation, helping the result gain more structure and detail. The third stage performs high-resolution refinement after latent upscaling, making the final output cleaner and more suitable for publishing.Compared with ordinary image-to-video workflows, this graph is more structured. A simple one-pass workflow may create motion but often struggles with identity drift, weak detail, inconsistent lighting, or unstable composition. This version uses staged sampling, manual sigma control, image conditioning, negative guidance, latent upscaling, and tiled decoding to improve control and final quality.This workflow is suitable for turning portraits, character designs, AI illustrations, cinematic stills, product visuals, anime scenes, fantasy concepts, and cover images into short video clips. It is especially useful for creators who want to make AI video previews, social media clips, MV fragments, Bilibili demonstrations, YouTube shorts, RunningHub showcases, or Civitai workflow examples from a single strong image.The final output is decoded through tiled VAE decoding and assembled into a playable video. This makes the workflow practical for both testing and actual content production.Main features:LTX 2.3 image-to-video workflowOne image to complete video clipOmniNFT + Relay-style prompt controlDistilled 1.1 model routeLTXVImgToVideoConditionOnly image guidanceLTXVPreprocess input preparationLTX2_NAG negative guidance supportUniversal negative prompt structureThree-stage rendering pipelineManualSigmas and SamplerCustomAdvanced controlLTXVLatentUpsampler high-resolution refinementTiled VAE decoding and final video outputSuggested workflow:Prepare one clean source image first. The subject should be clear, visually stable, and not too small in the frame. Load the image into the workflow, then write a prompt that describes the desired motion, camera behavior, lighting, atmosphere, and video style. Run a short test first to check whether the image identity is preserved and whether the motion direction is correct. If the video is too static, strengthen the motion description. If the subject changes too much, reduce aggressive prompt wording and keep the image guidance stronger. Once the first stage looks stable, continue through latent upscaling and final high-resolution refinement.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058326590685274113?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058326590685274113?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Video Extension OmniNFT + Relay Vertical Widening Workflow
Watch the full video first if you want to understand how this LTX 2.3 video extension workflow works in practice. The video shows how a vertical video can be expanded into a wider frame, why OmniNFT + Relay guidance matters, and how to launch the workflow online without rebuilding the full ComfyUI setup locally.This ComfyUI workflow is designed for LTX 2.3 video extension, vertical-to-wide frame expansion, and guided video outpainting. The main purpose of this workflow is to take a narrow or vertical video-style input and expand it into a wider composition while keeping the original subject, motion, and visual identity as stable as possible. Instead of simply stretching the image or cropping the video, this workflow uses LTX 2.3 generation to synthesize new side areas and create a more natural wide-frame result.The workflow is built around the LTX 2.3 distilled 1.1 route. It uses ltx-2.3-22b-dev-dare-merged-distilled-1.1.safetensors as the main checkpoint, Gemma-based text encoding, LTX Audio VAE, LTXVConditioning, LTXAddVideoICLoRAGuide, LTXVCropGuides, LTXVConcatAVLatent, LTXVSeparateAVLatent, ManualSigmas, CFGGuider, SamplerCustomAdvanced, VAEDecodeTiled, and VHS_VideoCombine. The graph also includes image resizing, color correction, image switching, and image concatenation nodes, which are important for comparing and assembling the expanded output.The key idea is guided extension. The source frame or reference image is first prepared through resizing and LTXVPreprocess. Then LTXAddVideoICLoRAGuide injects the visual guide into the LTX generation process, helping the model preserve the original content while expanding beyond the initial frame boundary. LTXVCropGuides helps manage the guided area so the model can focus on the extension region instead of freely changing the whole image.The workflow also uses audio-video latent logic. Empty video latent and empty audio latent are created, then combined through LTXVConcatAVLatent before sampling. After generation, LTXVSeparateAVLatent separates the video and audio latent streams again. This makes the workflow compatible with LTX 2.3 audio-video generation structure and final video output.Compared with ordinary video resizing, this workflow does not only change the canvas size. It generates new visual content for the expanded region. Compared with basic image outpainting, it works in a video pipeline, so it is more suitable for motion clips, MV fragments, portrait-to-landscape conversion, social media repurposing, and cinematic reframing. It is useful when you have a vertical clip but want to create a wider version for YouTube, Bilibili, horizontal previews, cover videos, or cinematic presentation.The final output is decoded through tiled VAE decoding, optionally adjusted through color correction, assembled through ImageConcanate, and exported through VHS_VideoCombine as an MP4 video. This makes the workflow practical for direct online testing and publishing.Main features:LTX 2.3 video extension workflowVertical-to-wide video expansionOmniNFT + Relay-style guided generationDistilled 1.1 model routeLTXAddVideoICLoRAGuide visual guidanceLTXVCropGuides guided area controlImageResizeKJv2 and LTXVPreprocess input preparationManualSigmas and SamplerCustomAdvanced samplingAV latent concatenation and separationColor correction and image switching supportImageConcanate for side-by-side assemblyVHS_VideoCombine MP4 exportSuggested workflow:Prepare a vertical or narrow source image/video frame first. Make sure the main subject is clear and not too close to the edge unless you intentionally want strong side expansion. Load the source into the workflow, check the resize and guide settings, then run a short test first. If the expanded area looks too weak, strengthen the guide or adjust the prompt. If the original subject changes too much, reduce aggressive prompt wording and keep the guide structure stable. Once the expansion looks natural, use the final video combine section to export the widened MP4 result.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058327114209910786?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058327114209910786?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Multi-Image Reference OmniNFT + Relay Video Fusion Workflow
Watch the full video first if you want to understand how this LTX 2.3 multi-image reference workflow works in practice. The video shows how multiple images are fused into one video generation pipeline, why OmniNFT + Relay control matters, and how to launch the workflow online without rebuilding a complex local ComfyUI environment.This ComfyUI workflow is designed for LTX 2.3 multi-image reference video generation, using OmniNFT, Relay-style prompt control, and the distilled 1.1 model route to create more controllable image-to-video results. The main purpose of this workflow is to let creators use several reference images at the same time instead of relying on only one starting frame. This makes it much more suitable for character consistency, multi-angle visual guidance, object continuity, environment reference, and cinematic video generation.The workflow is built around the LTX 2.3 distilled 1.1 video route. It uses a Gemma3-based LTX text encoder, LTX video VAE, LTX Audio VAE, LTXVConditioning, LTX2_NAG for stronger negative guidance, and LTXVAddGuideMulti for multi-image reference control. The workflow also uses ManualSigmas, CFGGuider, RandomNoise, SamplerCustomAdvanced, LTXVConcatAVLatent, LTXVSeparateAVLatent, LTXVLatentUpsampler, tiled VAE decoding, and CreateVideo output.The key node is LTXVAddGuideMulti. This node allows multiple images to act as visual guides across the video timeline. Each guide can be assigned a frame index and strength value, so the workflow can control when a reference image becomes important and how strongly it affects the output. One image can define the main character, another can define the scene, another can provide clothing or object details, and another can guide a later-frame direction.The workflow uses a three-stage rendering structure. The first stage focuses on initial composition and motion foundation. The second stage handles latent continuation and latent-space expansion. The third stage performs high-resolution refinement after the latent upscaler. This staged approach is more stable than forcing the whole video into one single pass.Compared with ordinary image-to-video workflows, this graph gives creators stronger control over visual continuity. A normal single-image workflow may struggle with stable identity, clothing consistency, background logic, or multi-reference storytelling. This multi-image workflow gives the model several visual anchors, making it better for character videos, cinematic shots, MV fragments, product-style video, and advanced LTX 2.3 demonstrations.The final output can be decoded through tiled VAE decoding, combined with audio through CreateVideo, and prepared for publishing or further editing. This makes the workflow useful not only for testing, but also for practical video production.Main features:LTX 2.3 multi-image reference video workflowOmniNFT + Relay-style multi-reference controlDistilled 1.1 model route for practical generationLTXVAddGuideMulti frame index and strength controlMultiple image references for character, scene, object, and style guidanceLTX2_NAG negative guidance supportThree-stage rendering structureManualSigmas and SamplerCustomAdvanced controlLTXVLatentUpsampler high-resolution refinementAudio-video latent concatenation and separationTiled VAE decodingCreateVideo final outputSuggested workflow:Prepare several clear reference images first. Use one image for the main character, one for the environment, one for clothing or object details, and one for a later-frame visual direction. Load them into the reference image inputs, then check the frame index and guide strength values inside LTXVAddGuideMulti. Start with a short test render to confirm whether the references are being fused correctly. If the output becomes chaotic, reduce guide strength or simplify the prompt. If a reference is too weak, increase its strength or move its frame index closer to the target moment. After the base composition is stable, continue into latent upscaling and final high-resolution refinement.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2058327091539701762?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1yRGj6XEaM/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.⚙️打开下方链接即可在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2058327091539701762?inviteCode=rh-v1111如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1yRGj6XEaM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX2.3 All in one [SFW / NSFW] - LTX Director + ID LoRA + ControlNet + Detailer + Upscaler
Thanks everyone for the Buzz ⚡ and the feedback 😊⚠️ Warning: Although this is not an especially advanced workflow, it is also not beginner-friendly. It uses custom nodes and, depending on the setup, may require basic knowledge of Python dependency management, virtual environments, the ComfyUI file system, and troubleshooting.If you are unfamiliar with these topics, please make a backup of your ComfyUI setup before installing new nodes or updating ComfyUI.For beginners, I generally recommend using the portable version, as it is easier to maintain and back up.It is also important to have a solid understanding of how to properly use and prompt the LTX2.3 model, including concepts such as guidance strengths, valid resolutions, reference image quality, and related settings. This is especially important when using the 10Eros model, as it is significantly less forgiving than the original model.This workflow supports 3 types of models currently:Standard LTX 2.3 distilledLTX 2.3 distilled GGUF10Eros💡 The models are self-contained. You can safely delete the entire group of whichever model you don't use without breaking the workflow. The remaining model groups will work independently without any additional changes needed.💡 Two workflow versions are included in the zip file:LTX23Legacy.json -> Legacy version for those whom like things "hands on" (No LTX Director node)LTX23_V3_LTXDirector.json -> Same features, but using LTX Director node.⚠️ COMFYUI SECURITY ERROR FIX: At the end of this post.🎬 This is a total revamp of the original workflow with the same features, but using LTX Director.You can find the author's tutorial here.This workflow is a modular and flexible text/image/audio-to-video generation system built in ComfyUI, designed to give full control over video creation using LTX-based models. It allows you to easily mix and match multiple generation modes such as text-to-video, image-to-video, lipsync, and fully guided animation by enabling or disabling grouped nodes.📝 Personal notes:The 10Eros model is better for NSFW content, whereas the standard model is better for SFW generations, although the body movement of the 10Eros model can be beneficial in some cases for SFW content too, but in general, use each model as I just said.Try to always use 2 phase sampling generations (Half res + 2x upscaler), this yields the best quality and character consistency, LTX is not good at all at preserving character ID, so don't make it worse by doing a single pass generation. The upscaler model adds extra detail and improves character consistency, that's why I recommend using it.However, using half res on resolutions under 480p will generate bad quality videos since, at that point, you would be creating a first pass video so small that the upscaler won't have much details to work with. So for 480p and below, one single pass video generation is the way to go. On an RTX 4090, you can generate 1440p short videos using half res + upscaler, however, I think the sweet spot between quality and speed is at 720p.Don't use the detailer when generating "Amateur look" videos, it adds a light layer of detail to the final result, and most of the time it will look too "polished" for a real amateur recording; amateur style videos look more real when they look low quality.Main featuresLTX Director supportGGUF supportSFW / NSFW supportNSFW prompt enhancerText, image, audio, and ControlNet-driven video generationLoRA support (character, style, and voice via ID LoRA)Custom or AI-generated audio with automatic syncingReference image + unlimited keyframes (FFLF animation control)ControlNet video guidance with hybrid reference supportHalf-res sampling + 2× upscaling for faster high-quality resultsLTX detailer for enhanced final outputCommon SetupsText to video:Timeline text segment.Image to video:Add an image segment to LTX Director, extend it to fill your video duration and input a segment prompt for that segment.Lipsync:Add an image segment to LTX Director, extend it to fill your video duration, input a segment prompt for that segment and add the audio file to the timeline. Enable custom audio in LTX director.Audio to video:Add a text segment to LTX Director, extend it to fill your video duration, input a segment prompt for that segment and add the audio file. Enable custom audio in LTX director.Character LoRA + voice reference:Add a text segment to LTX Director, extend it to fill your video duration and input a segment prompt for that segment. Add your character LoRA to the power LoRA loader, then activate ID LoRA and load a voice referenceVoice reference to video:Add a text segment to LTX Director, extend it to fill your video duration and input a segment prompt for that segment. Activate ID LoRA and load a voice referenceCharacter animation:Add an image segment to LTX Director, extend it to fill your video duration and input a segment prompt for that segment. Enable controlnet.First frame → last frame:Add two image segments to LTX Director, and shrink the second image segment (last frame), add your prompt in the first image segment and in the second image prompt write "."First → middle → last frame:Same thing as for FFLF, but adding an image segment in the middle and its corresponding prompt.Character animation with custom voice:Image segment + ID LoRA + ControlNetCOMFYUI SECURITY ERROR FIXIf you hit a security level error while trying to install some custom nodes from ComfyUI manager:You need to temporarily change the security level in your config.ini file.If you are using the portable version, the file is located at:ComfyUI/user/__manager/config.iniChange the value to:security_level = weakIMPORTANT:Do NOT forget to restore it back to its original value, you should restore it afterward for safety reasons.Perform a hard restart of ComfyUI (completely stop it, then start it again).Install the custom node(s) using ComfyUI manager and double-check that it is the official one.Restart ComfyUI and check if the node(s) work(s) properly.Once everything is working properly, restore your original security level in the config.ini file and restart ComfyUI one final time.Detailed instructions are contained in the workflow itself:Red nodes are instructions and useful notes.Yellow nodes are configurable elements you can adjust to your needs.LTX DIRECTOR VERSIONLEGACY VERSION
LTX2.3 I2v / T2v
ComfyUI WorkflowRequires: ComfyUI_Eclipse 3.5.xbased on the standard ltx 2.3 template slightly changedadded auto prompting (Smart LM Loader)Load Image of Eclipse because it's dom widget can be used in a SubGraphalso added some Get/Set nodes
LTX Director
The LTX Director node is a custom ComfyUI node designed to give more cinematic control over LTX video generation workflows. It helps create dynamic camera movement, scene continuity, and more directed motion inside a generated video sequence.It is especially useful for music videos, cinematics, and storytelling workflows where camera behavior, shot consistency, and scene flow are important.
Post Processing IMG/VIDEO Any
simple workflow to do the post processing for mainly videos.since i get many OOM. in some parts of the video generations. i split the post processing to avoid this problem.any content of this workflows dont require checkpoint models. only his own models like rife49 for interpolation. etc.etc.etc.v1.05 - requires now my own custom nodes to solve some issues from now onshirosaki33/shiroaudio: custom nodes for comfy, related to audiov1.055 - experimental toolsshirosaki33/shirotools: many utility tools made for different situationsthe image process is just a extra.you can do by single file or batch path in both.
Ltx 2.3 Fix Face Workflow ( 10Eros )
Hey Guys I have made a video on this Workflow Please check it out and all the models and links are in the descriptions of the video
For-Loop LTX 2.3 Long MV Auto Generation Workflow
This ComfyUI workflow is designed for LTX 2.3 long MV generation, audio-driven video creation, and for-loop style automatic music video production. Unlike a simple image-to-video workflow that only generates one short clip, this workflow focuses on longer video output by combining audio duration detection, automatic frame calculation, image-to-video conditioning, audio-video latent processing, latent upscaling, multi-stage sampling, and final video assembly.The workflow is built around LTX 2.3, using ltx-2.3-22b-dev as the main video model, Gemma 3 12B as the text encoder, LTX 2.3 spatial upscaler for latent enhancement, and motion/control LoRA support for stronger video consistency. It can take image input, audio input, prompts, and automatically calculate the number of frames needed for the video. The frame logic follows the LTX-compatible 8n+1 rule, helping users avoid frame-count errors when matching video duration to music or narration.A key part of this workflow is the automatic duration system. The audio duration is read, converted into frame length, and aligned with the required LTX frame structure. This makes the workflow more practical for MV production because users do not need to manually calculate every segment. The workflow also uses LTXVConditioning, LTXVImgToVideoConditionOnly, LTXVConcatAVLatent, and LTXVSeparateAVLatent to connect image guidance, audio-video latent logic, and video generation.The workflow is structured for long-form generation. Instead of forcing the whole MV into one single heavy render, it uses a staged process. The first stage creates the base motion and visual direction. Later stages can continue, refine, upscale, and improve the latent video result. This makes it easier to build longer music videos, character MVs, digital idol clips, cinematic visual loops, and stylized AI video segments.The workflow also includes LTXVLatentUpsampler for higher-quality output. This allows the video to be generated more efficiently at a manageable stage first, then enhanced later through latent upscaling and additional refinement. This is useful for balancing speed, quality, and VRAM usage.Final output is handled through VHS_VideoCombine, which combines the generated frames with the audio track into a finished MP4 video. This makes the workflow suitable for actual publishing, not just frame preview. It can be used for YouTube, Bilibili, RunningHub demonstrations, Civitai workflow showcases, social media clips, and AI music video experiments.Main features:- LTX 2.3 long MV generation workflow- Audio-driven video creation- Automatic audio duration detection- 8n+1 compatible frame calculation- Image-to-video conditioning- Multi-image input support- Audio-video latent processing- ManualSigmas and SamplerCustomAdvanced control- LTXVLatentUpsampler for quality improvement- Final MP4 output with audio- Suitable for long MV, music video, digital idol, character video, and AI visual storytellingSuggested workflow:Prepare your audio first, then prepare one or more reference images. Use a clear prompt describing the subject, scene, camera movement, lighting, mood, and MV atmosphere. Let the workflow calculate the frame count automatically. Start with a shorter test segment, confirm the motion and identity are stable, then use the loop-style structure to extend and refine the full MV.🎥 YouTube Video TutorialWant to know what this workflow actually does and how to start fast?This video explains what the tool is, how to launch the workflow instantly, and shares my core design logic — no local setup, no complicated environment.Everything starts directly on RunningHub, so you can experience it in action first.👉 YouTube Tutorial: https://youtu.be/eqTjKOURnF0Before you begin, I recommend watching the video thoroughly — getting the full context helps you understand the tool faster and avoid common detours.⚙️ RunningHub WorkflowTry the workflow online right now — no installation required.👉 Workflow: https://www.runninghub.ai/post/2055888057307549697?inviteCode=rh-v1111If the results meet your expectations, you can later deploy it locally for customization.🎁 Fan Benefits: Register to get 1000 points + daily login 100 points — enjoy 4090 performance and 48 GB super power!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you’re in the Asia-Pacific region, you can watch the video below to see the workflow demonstration and creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1CSLJ6vEMQ/☕ Support Me on Ko-fiIf you find my content helpful and want to support future creations, you can buy me a coffee ☕.Every bit of support helps me keep creating — just like a spark that can ignite a blazing flame.👉 Ko-fi: https://ko-fi.com/aiksk💼 Business ContactFor collaboration or inquiries, please contact aiksk95 on WeChat.🎥 YouTube 视频教程想了解这个工作流到底是怎样的工具，以及如何快速启动？视频主要介绍 工具定位、快速启动方法 和 我的构筑思路。我们会直接在 RunningHub 上进行演示，让你第一时间看到实际效果。👉 YouTube 教程： https://youtu.be/eqTjKOURnF0开始前建议尽量完整地观看视频 —— 把握整体思路会更快上手，也能少走常见弯路。⚙️ 在线体验工作流现在就可以在线体验，无需安装。👉 工作流： https://www.runninghub.ai/post/2055888057307549697?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1CSLJ6vEMQ/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX-2.3 Director (Distilled GGUF) workflows
OverviewThis workflow is based on LTX-2.3 Director nodes. It allows you to visually specify image ordering and keyframe timing. Special thanks to WhatDreamsCost for providing the powerful custom nodes.
IC Edit Subtitle & Watermark Removal Video Cleanup Workflow
This workflow is designed for IC Edit-style subtitle and watermark removal, built on an LTX 2.3 video restoration pipeline. Its main purpose is to take an existing video with hardcoded subtitles, captions, random AI text, logo overlays, signatures, platform watermarks, or semi-transparent marks, and reconstruct a cleaner video result while preserving the original motion, subject identity, camera movement, lighting, and scene composition.Unlike a simple blur, crop, mosaic, or overlay method, this workflow uses a generative restoration approach. It does not just cover the unwanted text area. Instead, it uses LTX 2.3, subtitle-removal IC LoRA, watermark-removal IC LoRA, video frame extraction, prompt-guided reconstruction, latent inpainting-style processing, tiled VAE decoding, and final video recombination to rebuild the hidden background more naturally.The workflow uses ltx-2.3-22b-dev-dare-ties-distilled-1.1 as the main model route, with LTXVAudioVAELoader, CheckpointLoaderSimple, LTXAVTextEncoderLoader, LTXVConditioning, LTXAddVideoICLoRAGuide, LTXVImgToVideoConditionOnly, VAEEncodeForInpaint, SamplerCustomAdvanced, LTXVCropGuides, VAEDecodeTiled, VHS_LoadVideo, and VHS_VideoCombine. It also loads two dedicated IC LoRA models: ltx2.3-ic-subtitles-remove-general-lora and ltx2.3-ic-watermark-remove-general-lora. This makes the workflow focused on cleanup and reconstruction rather than ordinary image-to-video generation.The positive prompt is specifically written for removal tasks. It asks the model to remove subtitles, captions, hardcoded text, AI-generated garbled text, platform watermarks, logo overlays, signatures, and semi-transparent marks. At the same time, it asks the model to restore the underlying image using surrounding visual context while preserving facial features, body shape, object boundaries, lighting, texture continuity, camera motion, and scene composition. This is the correct logic for video cleanup: erase the unwanted layer, but do not destroy the original scene.The negative prompt suppresses common restoration failures such as blur, oversaturation, pixelation, low resolution, grain, distortion, noise, compression artifacts, JPEG artifacts, glitches, watermark, text, logo, signature, copyright marks, subtitles, distorted sound, saturated sound, and loud audio artifacts. This helps reduce the chance that the workflow removes one text artifact but creates another.The workflow also keeps the audio and timing structure through the VideoHelperSuite route. VHS_LoadVideo extracts video frames and audio, while VHS_VideoCombine recombines the repaired frames with the original audio into a final MP4. This is important for real publishing use because many frame-only cleanup workflows lose the audio or break the original timing.This workflow is ideal for AI video cleanup, subtitle removal, watermark removal, logo removal, random text cleanup, hardcoded caption repair, social-media video restoration, AI-generated video polishing, RunningHub demos, Civitai previews, YouTube examples, and Bilibili workflow showcases. If you need a clean video result without visible subtitles, fake text, logo overlays, or watermark artifacts, this workflow provides a practical IC Edit-style restoration route.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2054546081706463234?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1za5y6FE7r/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2054546081706463234?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1za5y6FE7r/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
IC Edit HD Video Restoration & Enhancement Workflow
This workflow is designed for IC Edit-style high-definition video restoration and enhancement, built on an LTX 2.3 video upscale / repair pipeline. Its main purpose is to take an existing low-quality or compressed video, preserve the original composition and motion, and rebuild the final output with cleaner details, reduced artifacts, better texture, and a more stable high-definition look.Unlike a simple video upscaler or sharpening filter, this workflow is closer to a generative restoration pipeline. It does not only enlarge the frame or add artificial sharpness. Instead, it uses LTX 2.3, IC LoRA video upscale models, video frame extraction, prompt-guided enhancement, latent reconstruction, tiled VAE decoding, and final video recombination to improve the clip while keeping the original structure intact. This makes it useful when the goal is not to create a new video from scratch, but to repair and polish an existing result.The workflow uses ltx-2.3-22b-dev-dare-ties-distilled-1.1 as the core model route, with LTXVAudioVAELoader, CheckpointLoaderSimple, LTXAVTextEncoderLoader, LTXVConditioning, LTXAddVideoICLoRAGuide, LTXVImgToVideoConditionOnly, VAEEncodeForInpaint, SamplerCustomAdvanced, VAEDecodeTiled, VHS_LoadVideo, and VHS_VideoCombine. It also loads IC LoRA upscale models such as ltx2.3-ic-video-upscale-general and ltx2.3-video-upscale-v2, which shows that the workflow is specifically tuned for video quality restoration rather than normal image-to-video generation.A key part of this workflow is the positive enhancement prompt. The workflow asks the model to enhance the input video to clean high-definition quality, remove compression artifacts, noise, blur, jagged edges, color blocks, motion smearing, local dirt, and frame flickering, while reconstructing clearer facial details, skin texture, hair strands, clothing fabric, and background structure. At the same time, it explicitly preserves the original composition, character identity, motion rhythm, camera movement, lighting atmosphere, and color style. This is the correct direction for repair work: improve quality, but do not rewrite the video.The negative prompt is also practical. It suppresses blur, oversaturation, pixelation, low resolution, grain, distortion, noise, compression artifacts, JPEG artifacts, glitches, watermark, text, logo, signature, copyright marks, subtitles, distorted sound, saturated sound, and overly loud audio artifacts. This makes the workflow suitable for cleaning AI-generated clips, compressed social-media videos, rough preview renders, low-bitrate outputs, and unstable video drafts.The workflow also preserves audio and frame timing through VideoHelperSuite loading and combining logic. This matters because many repair workflows only process frames and lose the original audio or timing structure. Here, the video frames and audio route can be recombined into a final MP4, making the result more useful for direct publishing.This workflow is ideal for creators who want to improve AI video outputs before posting them on YouTube, Bilibili, RunningHub, Civitai, TikTok, or other platforms. It can be used for AI video cleanup, HD restoration, compression artifact removal, facial detail recovery, texture enhancement, frame stability improvement, and final “publish-ready” polishing. If you want to see how IC Edit HD repair, LTX 2.3 video upscale LoRAs, prompt-guided restoration, tiled decoding, and final video export work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2054546073984749570?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1za5y6FE7r/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2054546073984749570?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1za5y6FE7r/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Dual Digital Human | IC Edit No-Subtitle Dialogue Workflow
This workflow is designed for LTX 2.3 dual-person digital human dialogue generation, with IC Edit-style control and a strong focus on clean subtitle-free output. Its main purpose is to take a two-person reference image or character scene, generate a controlled dialogue-style video, and keep the final result clean without unwanted subtitles, fake captions, random text, watermark-like marks, overlays, or UI-style artifacts appearing on the screen.Compared with a single-person digital human workflow, this setup is more demanding because it needs to maintain two character identities at the same time. A good dual-person dialogue video must preserve left-right placement, facial consistency, clothing, body proportion, camera framing, background stability, and interaction logic. If the workflow is not controlled well, the two characters may swap positions, merge faces, duplicate body parts, drift away from the original image, or create random mouth movement that does not match the intended dialogue structure.The workflow uses LTX 2.3 as the main video generation backbone, with LTX video VAE, LTX audio VAE, image resizing, image-to-video conditioning, LTXVConditioning, LTXVImgToVideoConditionOnly, LTXVConcatAVLatent, LTXVSeparateAVLatent, SamplerCustomAdvanced, ManualSigmas, latent upscaling, tiled VAE decoding, audio decoding, CreateVideo, SaveVideo, and VRAM cleanup logic. This makes it a more complete production workflow rather than a simple one-pass image animation graph.The core generation design follows a staged rendering structure. The first stage builds the base motion, character presence, camera structure, and dialogue performance from the reference image and prompt conditioning. Later stages continue from the generated latent result with lower sigma values, refining motion stability, facial detail, clothing texture, background consistency, and final visual quality. This staged approach is especially useful for two-person digital human scenes because both subjects need to remain coherent across the full video.A key feature of this workflow is its dual-character control direction. The workflow is built for restrained dialogue performance rather than chaotic motion. The ideal output should show two people facing the camera or interacting naturally, with subtle head movement, mouth movement, facial expression changes, small hand gestures, and stable body posture. The workflow is not intended to create excessive action; it is designed to create clean, usable talking scenes for AI presenters, short drama dialogue, virtual hosts, product explanation, teaching videos, and social media content.The no-subtitle direction is another important selling point. Many AI video models may accidentally generate fake subtitles, random text, caption bars, logos, watermarks, or noisy symbols when the prompt contains speech or dialogue. This workflow is designed with clean-output restrictions to reduce those problems and make the generated dialogue video more suitable for direct publishing.The workflow also uses audio-duration handling, FPS control, frame-count logic, latent preparation, mask / latent routing, latent upscaling, tiled decoding, and final video export. These components help creators generate repeatable dual-person dialogue clips while managing memory pressure and maintaining final output quality.This workflow is ideal for creators who want to produce two-person AI dialogue videos, dual digital human presentations, character conversation clips, virtual interview scenes, AI short-drama dialogue, Bilibili / YouTube explainers, RunningHub demos, and Civitai workflow previews. If you want to see how LTX 2.3, IC Edit-style control, staged sampling, dual-character stability, no-subtitle restrictions, and final dialogue video export work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow OnlineWorkflow: https://www.runninghub.ai/post/2054479470995755009?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1za5y6FE7r/I will continue updating model resources on Quark Drive:https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流工作流： https://www.runninghub.ai/post/2054479470995755009?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1za5y6FE7r/我会在 夸克网盘 持续更新模型资源：https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
videoLab - LTX 2.3
*When copy+pasting the workflow Audio VAEs disconnect and Math Expression #74 has to be reconnected from it's port b to Frame Rate*Check out my other workflows for swap and upscale techniquesLTX 2.3 Super speaks — but now it speaks in your voice.Your Custom VoiceLTX audio merged inLip sync stabilityThis workflow combines voice cloning with LTX 2.3's native audio generation. One prompt produces two synchronized audio tracks — cloned dialogue and AI-generated background sound — merged and conditioned into the video latent.Your voice. Your script. LTX's world. One prompt.LTX 2.3 — AI Video Workflow (part 4 of Lab series)Four inputs. One cinematic video.Drop in an image, a voice sample, your scene description, and your character's dialog — the workflow takes it from there.✍️ Script-Driven Dialog Write exactly what you want your character to say in the Dialog Script node. Your words, your story — the workflow handles all the technical execution automatically.🎬 Scene Description Describe the scene, the mood, the environment. The AI reads this alongside your dialog and builds everything else — no prompt engineering knowledge needed.🤖 Fully Automatic Prompting Two local AI enhancers run silently in the background. One writes a complete structured video generation prompt from your scene and dialog. The other crafts precise voice delivery direction — pace, tone, emotion, texture — fed directly into the voice engine. You never write a prompt manually.🎙️ VoX Voice Cloning Provide any short voice sample and VoxCPM2 clones it. Your character speaks your script in that voice, with the AI-generated delivery direction shaping how it's performed — not just what is said, but how.🔁 Dual Voice Reinforcement The video and audio pipelines are locked together. The video prompt instructs the model on lipsync behavior while the audio direction drives matching emotional delivery through VoX — a unified performance from both sides simultaneously.⚙️ GGUF & Safetensors Compatible Run whichever model format your hardware prefers — GGUF quantized or full safetensors, both supported out of the box.
LTX-2.3 All-In-One workflow for RTX 3060 with 12 GB VRAM + 32 GB RAM
[edit:13.05.2026: Update version 4.4 (see version description).Small fixes to get back fast generations.Attention:If you struggle with node conflicts or you get errors while running the workflow, please have a look at my short Trouble Shooting Guide note in the wokflow first. Most importent is to update all components sucsessfully! ]Special thanks to:@ArcleinSK for investigation and solving the FLF issue, as well as forcing the First-Mid-Last Frame option and last but not least for charing fantastic knowlage.@boinobin730 for initialising, forcing and supporting this project in all kinds of matter, like providing links, running tests, sharing knowlage and inspiring diskussions.@Urabewe for publishing the original, perfectly running 12 GB VRAM LTX-2.3 workflows mainly used here in this workflow.Features:Simple to use all-In-One LTX-2 workflow with options for:Text to VideoImage to VideoFirst/Last Frame to VideoFisrt/Mid/Last Frame to VideoVideo to VideoText + Audio to VideoImage + Audio to VideoFirst/Last Frame + Audio to VideoFirst/Mid/Last Frame + Audio to Videoeasy switching between all options,all steps highly automated: no manual frame or width/hight calculations necessary,easy to set inputs by predefined sliders and aspeckt ratio inputs (no risk to set wrong frame counts or wrong width/hight values),completely automated resizing and cropping (if necessary) of your input images/videos.brilliant audio generation (speech/sound) with LTX-2.3.LTX-2.3 specifications:Workflow version v4.3 consistently follows the LTX-2.3 specifications for 16:9/9:16 aspect ratios, including automatic width/hight calculations, as well as automatic input image/video resizing/cropping.In addition you can simply choose now any other aspect ratios according to your needs while still getting the right values calculated for width/hight and automatic image/video resize/crop.Requirements:GPU with 12 GB VRAM (some users reported they got it running with 8 GB too),32 GB VRAM,Swap file size: 64 - 128 GB.Speed and video length:Runs very fast: 5 second (1280 x 864) Video: < 10 minutes.Generation of long high quality videos in one run possible: 10 - 20 seconds without any issues,Testrun: 30 second video (1024 x 704) tooks around 40 minutes without any OOM errors. Longer videos might be possible, but not tested yet.Important:This workflow is intended for advanced comfyui users who know how to install and operate the system and are able to resolve basic system errors themselves, like as node conflicts, or general system issues.About this workflow:This workflow is mainly based on the fantastic LTX-2.3 workflows of @Urabewe.As far as I know, those were the first workflows running LTX-2 with 12 GB VRAM. All credits goes to the original creator.My job was only to combine and organise the different workflows in a simple to use all-in-one design.
LTX 2.3 Video Control & HD Enhancement Workflow
This workflow is designed for LTX 2.3 video control and high-definition enhancement. Its main purpose is to take a video or image-guided video input, preserve the original motion structure, and enhance the final result through LTX 2.3 generation, control preprocessing, latent upscaling, audio-video latent routing, and tiled decoding. It is built for creators who want a cleaner, sharper, more stable LTX 2.3 video output instead of a rough low-resolution generation pass.The workflow uses LTX 2.3 as the main video generation backbone, with ltx-2.3-22b-dev_transformer_only_fp8_scaled as the core model route. It also includes Gemma-style LTX text encoding, LTX23 video VAE, LTX23 audio VAE, LTXVPreprocess, EmptyLTXVLatentVideo, LTXVEmptyLatentAudio, LTXVConcatAVLatent, LTXVSeparateAVLatent, SamplerCustomAdvanced, LTXVLatentUpsampler, VAEDecodeTiled, LTXVAudioVAEDecode, and final video output logic. This makes the graph more advanced than a simple image-to-video or video-to-video workflow because it is structured around both control and enhancement.A major strength of this workflow is its video-control preparation section. The graph includes VideoHelperSuite video information reading, frame rate extraction, frame count handling, image resizing, and LTX preprocessing. It also includes optional control preprocessing routes such as DepthCrafter, Canny edge extraction, and DW pose preprocessing. These control modules are useful when the creator wants the generated result to follow the source video’s structure, depth, body movement, edge layout, or camera rhythm more closely.The workflow is also designed for HD improvement. Instead of decoding only the first latent result, it uses LTXVLatentUpsampler and additional refinement sampling stages to push the video toward a higher-quality output. This helps improve detail density, texture clarity, subject sharpness, and final frame polish. The tiled VAE decoding route is important here because high-resolution video decoding can easily become memory-heavy or unstable. Tiled decoding allows the workflow to decode large frames more safely and cleanly.The audio-video route is another practical part of the graph. Audio latent and video latent are connected, separated, decoded, and preserved through the generation process. This makes the workflow suitable for real video production rather than silent visual testing. For AI short videos, character clips, music-driven scenes, product demos, cinematic loops, and social media publishing, keeping audio and visual timing together is a major advantage.This workflow is especially useful for creators who already have a video structure and want to improve it with LTX 2.3 rather than starting completely from zero. It can be used for video enhancement, controlled video regeneration, pose-guided clips, depth-guided cinematic shots, AI video cleanup, higher-resolution remakes, and RunningHub / Civitai workflow demonstrations.If you want to see how the control video, LTX 2.3 model route, Depth / Canny / Pose preprocessing, latent upscaling, tiled decoding, and final HD video export are connected, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2040812593555709953?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1gaSfBgEqz/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2040812593555709953?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1gaSfBgEqz/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
Anima Preview3 | Image-to-Image Anime Refinement Workflow
This workflow is designed for Anima Preview3 image-to-image generation, focusing on controlled anime-style transformation from an existing reference image. Its main purpose is to let creators upload a source image, guide it with a text prompt, and generate a cleaner Anima Preview3 result while still preserving the basic structure, pose, composition, and subject direction of the original picture.The workflow uses anima-preview3-base.safetensors as the main generation model, qwen_3_06b_base.safetensors as the text encoder, and qwen_image_vae.safetensors as the VAE. This creates a compact Anima Preview3 img2img pipeline where the input image is first resized, encoded into latent space, edited through the sampler, decoded back into an image, and then previewed or saved. Compared with a pure text-to-image workflow, this setup gives creators a stronger visual anchor because the model is not starting from an empty latent. It starts from the uploaded image and modifies it according to the prompt.The image preparation section is simple and practical. The source image is loaded through LoadImage, then passed into image_scale_pixel_v2, with the total pixel target set around 1 megapixel and alignment set to 64. This helps normalize the image into a model-friendly size before VAE encoding. The workflow is therefore useful for taking an existing AI draft, sketch, screenshot, character image, animal image, or rough composition and pushing it into a more polished Anima Preview3 style.The workflow then uses VAEEncode to convert the scaled image into latent space. This is the key difference from text-to-image. In text-to-image, the model creates everything from noise. In image-to-image, the original image becomes the base latent, so the final result can keep more of the original layout. This makes it especially useful for redraws, style refinement, concept cleanup, character reinterpretation, and controlled anime transformation.The sampling stage uses ClownsharKSampler_Beta with a 30-step setup, beta57 scheduler, linear/euler sampler route, CFG around 3, and denoise around 0.73. That denoise value is important: it is strong enough to visibly transform the image, but still keeps the source image as a meaningful reference. Lower denoise would preserve the original more strictly; higher denoise would push the result closer to a full redraw.The example prompt is simple: “masterpiece, best quality, score_7, highres, safe, 1 dog is running.” This makes the workflow a clean baseline for testing how Anima Preview3 handles image-guided transformation. Users can replace it with character prompts, action prompts, environment prompts, clothing descriptions, lighting instructions, or style directions depending on the image they upload.The negative prompt suppresses common generation problems such as worst quality, low quality, low-score output, artist names, blur, extra fingers, bad hands, bad anatomy, malformed limbs, duplicated elements, cropped bodies, deformed faces, poorly drawn eyes, text, and watermark artifacts. This keeps the workflow practical for Civitai previews, RunningHub demos, cover images, prompt experiments, and daily anime-style image production.This workflow is ideal for creators who want a clean Anima Preview3 image-to-image setup without heavy extra modules. It is suitable for anime redraws, reference-based generation, character refinement, animal motion reinterpretation, concept art cleanup, and visual style testing. If you want to see how the source image, Anima Preview3 model, Qwen text encoder, Qwen VAE, denoise control, and final img2img output work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2042205599336763393?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1MEQbBsE5z/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2042205599336763393?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1MEQbBsE5z/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
VideoFlow - LTX 2.3 All-in-One T2V / I2V / A2V / Stable Character Voice, Wan 2.2/2.1 I2V workflow
Update 2026-05-13: LTX 2.3 All-in-One v3.0 workflow publishedVideoFlow LTX 2.3 distilled 1.1 All-in-One v3.0New Features:Text-to-Video Support with a pre-configured set of LoRAs for creating photorealistic videos.Image-to-Video Support for both first and last frame.Optional Audio Integration (audio-to-video) with the ability to extract voice from recordings (file or recorded clip) to remove background noise.Consistent Character Voice through voice cloning with just a 5-second reference audio (file or recorded clip).Video Filters for adjusting brightness, contrast, saturation, sharpness, blur, and enhancing edges and details.Film grain for a cinematic or analog effect.50fps Support via frame interpolation.Improvements:Now using LTX 2.3 distilled 1.1 resulting in better emotions, movements and audio.Faster, less memory-intensive color correction.More explanations and guidance integrated.Fixes:Audio and video are now always perfectly in sync.Resolution input (video dimensions) for image-to-video generation now works properly.Update 2026-04-14: LTX 2.3 I2V workflow updatedVideoFlow LTX 2.3 distilled I2V v2.0VideoFlow 2.0 is here, bringing major performance upgrades, better quality, and more flexibility to your workflow.Key Improvements:Much Faster Generation: Thanks to improved samplers and schedulers, videos generate approx. twice as fast compared to Version 1.0.Higher Quality Output: Despite the speed boost, image quality, audio quality, and prompt adherence have all been significantly improved.Flexible Model Support: You can now freely choose between multiple model types:CheckpointGGUF UNetDiffusion modelOptimized for Low VRAM Systems: With GGUF support, VideoFlow now runs much more efficiently on systems with limited GPU memory.Optional Sampler Preview: Disable the sampler preview to further reduce generation time.Improved Usability: Additional guidance and hint texts help you get the most out of the workflow.Update 2026-03-15: LTX 2.3 I2V workflow addedVideoFlow LTX 2.3 distilled I2V v1.0This workflow provides an easy-to-use image-to-video solution for LTX 2.3, designed to work seamlessly with the distilled LoRA model. It focuses on high-quality, realistic output, with the first-stage scheduler's sigma values finely optimized for best performance.Subgraphs are used to keep the main workflow streamlined and easy to navigate. A live preview is displayed during generation, allowing you to monitor progress and stop the process early if desired. Additionally, the first-stage video can be decoded for quick previewing. This feature lets you watch a lower-resolution version of the final video and cancel immediately if the result doesn’t meet expectations.As the distilled LoRA already delivers impressive quality in the first stage, you can skip the second stage entirely if your hardware has limited performance. An optional color-correction node is included to compensate for LTX’s tendency to introduce subtle color and lighting shifts, ensuring consistent visual quality.Update 2025-08-24: Wan 2.2 I2V workflow addedVideoFlow Wan 2.2 I2V v1.0VideoFlow is now fully optimized for Wan 2.2. It supports resolutions from 480p up to 720p, with the option to upscale smoothly to 1440p at 32fps. The process is accelerated by integrating Lightning LoRA during the final two-thirds of the generation steps, ensuring faster results without compromising quality. Importantly, Lightning LoRA does not influence the initial generation steps, preserving natural and fluid movements throughout the video. SageAttention with Triton is supported but not required. Instructions on how to set up and use the workflow are included within the workflow itself.VideoFlow Wan 2.1 I2V v1.0This image-to-video workflow is designed to generate smooth, realistic videos at 32 fps with a strong emphasis on fast, high-resolution output. At least 16 GB of VRAM is recommended for optimal performance. For additional speed improvements, you may also install SageAttention and Triton, though these are optional.It's fast 🚀!Sample videos were rendered at 768 × 1152 resolution and 16 fps, consisting of 81 frames, each video taking about 6 minutes to generate. The upscaling and frame interpolation to 1536 × 2304 resolution and 32 fps took approximately another 6 minutes on an RTX 4080 with 16 GB VRAM. Lower resolutions render even faster.Key configuration for the sample videos:Video model: Wan2.1 SkyReels V2 I2V 14B 720PLoRA: Lightx2vSteps: 4Sampler: dpmpp_sde_gpuScheduler: beta💡Comprehensive usage details and instructions are provided within the workflow itself.Sample images for input were created with my PhotoFlow workflow.The download of the workflow contains all sample videos, including the input image with its own workflow, the initial generated video and its upscaled counterpart, allowing for convenient side-by-side comparison.Leave a 👍 if you like the workflow 🙂.
JoyAI Image vs Qwen 2511 | Single-Image Editing Comparison Workflow
This workflow is designed for comparing JoyAI-Image single-image editing with Qwen Image Edit 2511. Its main purpose is to help creators test how different image-editing routes handle the same or similar visual editing task, especially when the goal is not only to change an image, but also to preserve identity, structure, lighting, texture, and final visual quality.The workflow contains a JoyAI-Image editing branch and a Qwen Image Edit 2511 comparison branch. The JoyAI branch uses JoyAI-Image-Und-merger_bf16 as the visual understanding / CLIP route, joy_image_transformer as the main image transformer, Wan2.1_VAE.pth as the VAE, JoyAI_Image_ENCODER for image-and-prompt conditioning, JoyAI_Image_LATENTS for latent preparation, JoyAI_Image_SM_KSampler for sampling, and JoyAI_Vae_Decoder for final decoding. This route is focused on direct single-image instruction editing: upload one image, write the edit command, and let the model transform the image while keeping the important visual identity.The Qwen 2511 branch uses qwen_image_edit_2511_bf16 with Qwen Image Edit 2511 Lightning 4-step LoRA, qwen_2.5_vl_7b_fp8_scaled as the visual-language encoder, qwen_image_vae, TextEncodeQwenImageEditPlus, reference-latent method nodes, KSampler, and tiled VAE decoding. This branch is useful for comparing how Qwen 2511 handles detail, lighting, identity retention, reference-image understanding, and speed under a lightweight accelerated setup.The workflow is especially useful for creators who want to answer practical editing questions: Which route keeps the original character identity better? Which one follows text instructions more directly? Which one changes the background more naturally? Which one produces better lighting and shadow integration? Which one is faster for batch testing? Which one is better for Civitai preview images, RunningHub demos, social media covers, or local production?The JoyAI example focuses on a blue-haired cyber mechanical girl being transferred into a bright grassland scene while preserving her facial features, twin-tail hairstyle, mechanical ear devices, neck structure, and metallic body. The prompt also asks for realistic outdoor lighting, grassland atmosphere, natural shadows, metal highlights, and reduced cutout feeling. This makes it a strong test for identity preservation and scene replacement.The Qwen 2511 side includes a portrait-style light-and-shadow enhancement direction, asking for clearer facial texture, richer lighting, realistic skin detail, studio-style portrait quality, and stronger subject separation. This makes it suitable for testing whether Qwen 2511 performs better at photographic refinement, lighting enhancement, and controlled image polishing.This workflow is ideal for creators who want to compare JoyAI-Image and Qwen Image Edit 2511 in one editing experiment. If you want to see how the two editing routes behave under real workflow conditions, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2043308481817616386?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1MEQbBsE5z/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2043308481817616386?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1MEQbBsE5z/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
VBVR Digital Human | Stable Talking Avatar Workflow
This workflow is designed for VBVR digital human video generation, focusing on stable talking-avatar animation from a single reference image and an audio input. Its main purpose is to help creators generate a more controlled digital human result where the character keeps the same face, framing, camera angle, clothing, and background while performing natural speaking motion.The workflow is built around an LTX 2.3 video generation pipeline with VBVR I2V LoRA enhancement, LTX audio / video latent routing, Gemma-style text encoding, LTX video VAE, LTX audio VAE, NAG enhancement, IC LoRA motion-track control, spatial latent upscaling, multi-stage sampling, tiled decoding, and final video export. Compared with a basic image-to-video workflow, this setup is more suitable for talking-avatar production because it combines visual anchoring, audio routing, and controlled motion guidance in one graph.The core idea is simple but important: keep the shot stable and make the woman speak. In digital human generation, the biggest problem is often not whether the image can move, but whether it moves too much. A weak workflow may change the face, zoom the camera out, alter the clothing, deform the mouth, create extra hands, or shift the background. This workflow is designed to reduce those problems by keeping the camera and scene steady while concentrating motion on the face, mouth, head, and subtle body performance.VBVR is used here as an image-to-video consistency and motion-control booster. It helps the model follow the source image more closely and reduces random drift during generation. This is especially important for digital human videos because the first frame usually defines the person’s identity. If the generated video loses that identity after a few seconds, the result becomes unusable for avatar content, product narration, AI presenters, or character dialogue.The workflow also includes an audio latent route. The audio is encoded through LTXVAudioVAEEncode, connected into the audio/video latent structure, and later separated and decoded for final output. This makes the workflow more than a silent animation setup. It is designed for speaking-person videos where the final result needs both visual motion and usable audio-video export.Another important part is the use of NAG and IC LoRA motion-track control. NAG helps stabilize generation guidance, while the motion-control LoRA helps reduce uncontrolled movement. Together, they make the video more suitable for restrained digital human performance: stable eyes, soft head movement, natural mouth motion, minimal body drift, and consistent framing.The pipeline uses several sampling and refinement stages. The first stage builds the base talking video from the image, prompt, and audio latent. Later stages use latent upscaling and additional sampler passes to improve texture, detail, and final quality. This helps the output look more polished for Civitai previews, RunningHub demos, YouTube tutorials, Bilibili showcases, and social media publishing.This workflow is ideal for AI presenters, talking avatars, virtual hosts, product explainers, character narration, short-form dialogue videos, and digital human testing with LTX / VBVR. If you want to see how VBVR, audio conditioning, LTX 2.3 staged sampling, NAG guidance, and motion-control LoRA work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2043983796604768258/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1PQQuBcEd5/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2043983796604768258/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1PQQuBcEd5/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
WAI-ANIMA Image-to-Image | Controlled Anime Style Refinement Workflow
This workflow is designed for WAI-ANIMA image-to-image generation and controlled anime-style refinement. Its main purpose is to take an existing image as the visual foundation, then use the WAI-ANIMA model pipeline to redraw, enhance, stylize, or push the image toward a cleaner anime key visual while still keeping the original composition as an anchor.Compared with a pure text-to-image workflow, this setup gives creators more control because the input image is not discarded. The uploaded image is loaded into the workflow, resized through a pixel-scaling node, encoded into latent space through the Qwen image VAE, and then processed through the WAI-ANIMA generation model. This means the final result is guided by both the source image and the written prompt, making it useful for style conversion, anime redraws, character enhancement, composition preservation, and prompt-based visual refinement.The workflow uses waiANIMA_v10 as the main UNet model, qwen_3_06b_base as the text encoder, and qwen_image_vae as the VAE. This combination is aimed at anime-style image generation with stronger prompt understanding and cleaner visual rendering. The workflow is compact, direct, and easy to modify, making it suitable for creators who want a simple but practical Anima image-to-image setup instead of a large multi-stage graph.The input image is scaled to around a 1-megapixel working size through image_scale_pixel_v2. This helps normalize the image before it enters the latent process, avoiding extremely small or oversized inputs that may reduce stability. After that, VAEEncode converts the image into latent space, allowing the sampler to modify the image with controlled denoise rather than generating from a blank latent.The sampler section uses ClownsharKSampler_Beta with a 30-step beta-style sampling route, CFG control, and a moderate denoise setting. This is important for image-to-image work. If denoise is too low, the result may barely change. If denoise is too high, the original structure can be lost. The setup here is meant to balance source-image preservation with visible anime-style transformation.The positive prompt in the workflow is built around a high-quality anime key visual: adult anime beauty, Anima style, fantasy atmosphere, cinematic composition, detailed clothing, dramatic scale, and polished illustration finish. The negative prompt suppresses common quality problems such as low quality, bad hands, malformed limbs, duplicate elements, cropped bodies, deformed faces, poorly drawn eyes, text, and watermark artifacts. This makes the workflow suitable for polished anime illustrations rather than rough experimental outputs.This workflow is useful for anime character redraws, image stylization, reference-based illustration enhancement, fantasy key visual creation, Civitai preview images, RunningHub demos, social media covers, and quick visual testing. It is especially practical when you already have a rough image, AI draft, screenshot, or reference composition, and you want WAI-ANIMA to reinterpret it into a cleaner anime-style result.If you want to see how the source image, WAI-ANIMA model, Qwen text encoder, Qwen VAE, controlled denoise, and final image output work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2046194201804677121/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1q7drB7Ecp/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2046194201804677121/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1q7drB7Ecp/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Distill 1.1 + VBVR 240K | High-Probability Digital Human Workflow
VThis workflow is designed for high-probability LTX 2.3 digital human generation, built around LTX 2.3 Distill 1.1 and VBVR 240K enhancement. Its main purpose is to create a more reliable image-to-video talking-person or digital-avatar result, where the character keeps a stable identity, controlled facial motion, cleaner body movement, and stronger final visual quality.The workflow uses an LTX 2.3 video generation structure with ltx-2.3-22b-distilled-1.1, distilled LoRA support, VBVR-style image-to-video enhancement, Gemma-based LTX text encoding, LTX video VAE, LTX audio VAE routing, NAG enhancement, IC LoRA motion-track control, spatial latent upscaling, custom sampling, tiled decoding, and final video export. This makes it more production-oriented than a simple first-frame animation workflow.The core advantage of this setup is probability and stability. Digital human generation is not only about making a still image move. The model must preserve the face, avoid identity drift, maintain the original clothing and composition, keep the speaking performance believable, and avoid random body motion. This workflow is designed to improve those weak points by using the input image as a strong visual anchor, then reinforcing the generation through LTXVImgToVideoConditionOnly, LTXVPreprocess, audio/video latent routing, and multiple refinement stages.The workflow includes a dedicated audio path. Audio can be encoded through LTXVAudioVAEEncode and connected into the video latent process, allowing the output to behave more like a digital human video rather than a silent image animation. This is useful for AI presenters, talking avatars, product explanation videos, character narration, short drama dialogue, virtual influencer clips, and commercial-style social media content.NAG enhancement is another important part of the workflow. It helps strengthen generation control and reduce unwanted drift during sampling. For digital human videos, this is especially useful because even small errors in the face, mouth, hands, or camera motion can make the result feel unstable. The workflow also uses a motion-track control LoRA to guide the movement more deliberately, helping the character perform with more controlled motion instead of random animation.The pipeline is also staged for better final quality. It first builds the base video from the input image and conditioning. Then it uses latent upscaling and additional sampler passes to improve detail, texture, and visual polish. This staged approach helps the result look less like a rough preview and more like a usable output for Civitai previews, RunningHub demos, YouTube tutorials, Bilibili examples, and real production testing.This workflow is ideal for creators who want a stronger LTX 2.3 digital human solution with better consistency, better motion reliability, and a higher chance of usable results from one setup. If you want to see how Distill 1.1, VBVR 240K, audio conditioning, NAG, motion-control LoRA, and staged refinement work together, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2047657281310953473/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1x6oLB8E1d/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2047657281310953473/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1x6oLB8E1d/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
Prompt Relay Director | LTX 2.3 Dual-Character LoRA Dialogue Workflow
This workflow is designed as a Prompt Relay director for LTX 2.3 storytelling, dual-character dialogue, and dual-character LoRA prompt formatting. Its main purpose is not to generate the video directly, but to help creators turn messy ideas, rough Chinese notes, images, storyboard grids, short dialogue drafts, or incomplete scene concepts into a clean LTX 2.3 Prompt Relay format using only two fields: global_prompt and local_prompts.The workflow uses an LLM API node as the prompt director. It can analyze uploaded images, multiple images, rough prompts, short Chinese phrases, dialogue ideas, style requests, duration requirements, no-dialogue instructions, and storyboard-style references. After analyzing the input, it restructures everything into a usable Prompt Relay prompt that is much easier to paste into an LTX 2.3 video workflow.The core value of this workflow is standardization. Many LTX 2.3 video failures are not caused by the model alone, but by prompts that mix global rules, timeline events, dialogue, camera instructions, sound design, and negative restrictions in the wrong place. This workflow separates those elements clearly. The global_prompt locks the full-video anchors: scene, character identities, left-right placement, costumes, props, lighting, camera baseline, sound atmosphere, dialogue status, and stability restrictions. The local_prompts only describe time-specific events: actions, gestures, reactions, camera movement, object changes, dialogue lines, and timed sound effects.This is especially useful for dual-character LoRA scenes. Two-person video generation is difficult because characters may swap sides, merge, change style, lose identity, or produce chaotic body movement. This workflow forces the prompt to lock left-right positions, stable identities, stable character styles, natural eye contact, controlled gestures, and medium two-shot camera logic. It also adds restrictions such as no face swapping, no duplicated characters, no extra people, no wrong mouth movement, no mismatched lip sync, no subtitles, no screen text, no watermark, no jump cuts, and no random costume or style changes.The workflow also handles dialogue intelligently. If the user provides dialogue, the prompt director shortens and rewrites it into natural Mandarin lines suitable for lip sync. If the user does not ask for dialogue, it does not force speaking. If the user explicitly requests silence, roaring, breathing, sound effects, or no dialogue, the workflow generates a no-dialogue version with action, ambience, and sound effects instead.The timing rules are also built into the workflow. A 5-second scene is divided into four segments, a 10-second scene into four larger segments, and a 20-second scene into five segments. This makes the final result easier to control because each segment has one main event instead of a chaotic pile of actions.This workflow is useful for AI short drama prompts, two-character dialogue scenes, virtual influencer clips, fantasy conversations, anime/live-action mixed scenes, creature interaction scenes, and LTX 2.3 Prompt Relay testing. If you want to see how this prompt director converts rough ideas into clean global_prompt and local_prompts fields, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2048717560673214466/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DT9zBbEZu/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2048717560673214466/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DT9zBbEZu/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Subtitle Remover | AI Video Cleanup 0.5 Workflow
This workflow is designed for LTX 2.3 video subtitle removal and visual text cleanup, built as a reference 0.5 version for repairing unwanted text pollution inside AI-generated or authorized video materials. Its main purpose is to help creators remove subtitles, captions, random text artifacts, watermark-like overlays, and other unwanted visual marks from a video while keeping the original scene, motion, lighting, and unmasked areas as stable as possible.Unlike a simple crop, blur, or cover-up method, this workflow is based on mask-guided video repair. The unwanted subtitle or text region is isolated through a mask pipeline, then the model reconstructs the damaged area using the surrounding visual context. This makes the result more natural because the repaired area is regenerated to match the original background, rather than being hidden by a flat patch or blurred block.The workflow uses an LTX 2.3 repair route with video VAE, audio VAE, LTX conditioning, custom sampling, mask processing, and audio-video export logic. It includes structured SetNode / GetNode routing for base model, video VAE, audio VAE, CLIP, FPS, final masks, audio, and resolution management. This makes the graph more modular and easier to reuse in a production environment, especially when the user needs to repeatedly process different videos with similar subtitle or text-contamination problems.A key part of this workflow is the mask preparation section. The workflow includes mask handling tools such as BlockifyMask, final mask routing, and latent noise mask logic. This matters because video subtitle repair depends heavily on the mask quality. If the mask is too small, the text may remain. If the mask is too large, the model may unnecessarily change clean background areas. A good mask should cover the unwanted text fully while preserving enough surrounding context for the model to rebuild the background naturally.The workflow also keeps the audio route in the graph. Audio can be carried through the pipeline and reattached to the final output, which makes the workflow more practical for actual video repair instead of isolated frame testing. The graph includes audio retrieval, audio trimming / duration logic, LTXVAudioVAEEncode, LTXVConcatAVLatent, and final video creation. This allows the repaired result to remain usable as a complete video output.The sampling route uses LTX 2.3 video conditioning, CFG guidance, manual sigma control, SamplerCustomAdvanced, latent conditioning, and final decode / export logic. This gives the workflow stronger control over reconstruction compared with a one-click repair pass. The goal is to keep the unmasked regions unchanged while letting the subtitle region be regenerated in a visually consistent way.This workflow is useful for cleaning AI-generated videos, removing accidental prompt text, fixing subtitle contamination, repairing unwanted overlay captions, restoring damaged areas in authorized footage, and preparing cleaner results for Civitai, RunningHub, YouTube, and Bilibili publishing. Since this is marked as a reference 0.5 version, it is best understood as a practical test workflow: useful for learning the subtitle-removal structure, mask logic, and LTX 2.3 repair process before building a more polished production version.If you want to see how the mask is prepared, how the LTX 2.3 repair route handles subtitle regions, and how the final cleaned video is exported with audio, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2048741719289630722/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DT9zBbEZu/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2048741719289630722/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DT9zBbEZu/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Dual-Character Dialogue | Enhanced Cinematic I2V Workflow
This workflow is designed for LTX 2.3 dual-character dialogue video generation, focusing on stable two-person interaction, controlled left-right positioning, stronger motion continuity, and more polished final video quality. Its main purpose is to help creators turn a single reference image into a cinematic two-character conversation shot, where both characters remain visually consistent and the scene keeps a coherent dialogue rhythm instead of drifting into random motion.The workflow uses LTX 2.3 as the main video generation backbone, with ltx-2.3-22b-distilled-1.1 as the core checkpoint route. It also uses Gemma-style LTX text encoding, LTX audio VAE routing, LTX video VAE decoding, image-to-video conditioning, NAG enhancement, VBVR I2V LoRA support, latent upscaling, multi-stage sampling, and final video export. This makes it more advanced than a basic I2V workflow because it is built specifically for controlled character interaction rather than simple first-frame animation.The core strength of this workflow is two-character stability. In many AI video workflows, two-person scenes are difficult because characters may swap positions, merge into each other, lose identity, change clothing, or break the left-right relationship. This workflow is designed to reduce those problems by using the input image as a strong visual anchor, then reinforcing the generation with LTXVImgToVideoConditionOnly, LTXVConditioning, and a structured prompt. The negative prompt also directly suppresses subtitles, scene cuts, glitches, warping, extra limbs, extra hands, static frames, low-quality motion, and unwanted transitions.Another important feature is the VBVR I2V LoRA route. The workflow loads an LTX 2.3 VBVR I2V LoRA, which helps strengthen image-to-video behavior, motion consistency, and prompt adherence. This is especially useful for dialogue-style videos, where small gestures, facial direction, eye contact, and body positioning matter more than large chaotic movement.The workflow also includes NAG enhancement. NAG is used to improve guidance stability and reduce generation drift during sampling. For dual-character dialogue scenes, this matters because the video must preserve not only the scene style, but also the relationship between the two characters. The left character should remain on the left, the right character should remain on the right, and both should continue facing or reacting to each other in a believable way.The generation pipeline is also multi-stage. It first builds the initial LTX video latent from the input image, audio latent structure, prompt conditioning, and sampler route. Then it uses LTXVLatentUpsampler and additional refinement sampling stages to improve detail, texture, and visual polish. Instead of exporting a rough first pass, the workflow pushes the video through several controlled refinement steps, giving the final result a cleaner and more cinematic finish.This workflow is suitable for AI short dramas, anime-style character dialogue, fantasy conversation scenes, virtual influencer interactions, two-person storytelling, product dialogue clips, roleplay videos, YouTube demos, Bilibili tutorials, RunningHub publishing, and Civitai workflow showcases. It is especially useful when you want a two-character shot that feels like a continuous scene rather than disconnected AI motion.If you want to see how the input image is prepared, how VBVR and NAG improve two-person stability, how the three-stage refinement route works, and how the final enhanced dialogue video is exported, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2048727968108781570/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DT9zBbEZu/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2048727968108781570/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DT9zBbEZu/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Text Artifact Remover | AI Video Cleanup Workflow
This workflow is designed for LTX 2.3 AI video text artifact cleanup, focusing on repairing unwanted subtitles, random AI-generated letters, watermark-like text pollution, overlay captions, ghost text, logo artifacts, and other visual contamination inside video frames. Its main purpose is to help creators clean their own generated videos or authorized materials by reconstructing the damaged area instead of simply blurring, cropping, or covering it.The workflow uses an LTX 2.3 video inpainting route based on a GGUF LTX 2.3 distilled model, LTX23 video VAE, LTX23 audio VAE, Gemma-style text conditioning, custom sampler control, and LoRA-assisted repair. The key repair direction is built around LTX 2.3 Edit Anything and inpaint-style LoRA logic, allowing the model to understand that the masked area should be regenerated while the unmasked region should remain unchanged.The core prompt is very direct: remove the subtitle, watermark, or text inside the masked area, reconstruct the occluded background naturally and seamlessly, and keep the original scene, camera angle, lighting, motion, composition, and all unmasked regions unchanged. This is exactly what makes the workflow useful for AI video cleanup. It is not trying to redesign the whole shot. It is trying to surgically repair the polluted area.The negative prompt is also targeted for this use case. It suppresses subtitles, captions, text, Chinese subtitles, watermarks, logos, overlay text, random letters, unreadable text, ghost text, flicker, color shift, inconsistent background, blurry patches, and duplicated edges. These negative controls are important because text-removal workflows often fail by leaving behind soft stains, repeated edges, or new unreadable letters. This setup tries to reduce those artifacts during regeneration.The workflow also contains an audio latent route, using LTXVAudioVAEEncode and LTXVConcatAVLatent. Even when the repair task is mainly visual, keeping the LTX audio / video latent structure makes the pipeline more suitable for actual video production. The video latent is sampled through SamplerCustomAdvanced, then separated, cropped, decoded, and prepared for output. This makes it closer to a real repair workflow rather than a single-frame test.This setup is useful for fixing AI-generated video mistakes, removing accidental prompt text, cleaning subtitle pollution, repairing logo-like artifacts in authorized footage, restoring damaged visual regions, and preparing cleaner examples for Civitai, RunningHub, YouTube, and Bilibili. The most important usage rule is mask quality: the mask should cover the unwanted text area cleanly while leaving enough surrounding context for the model to reconstruct the background.If you want to see how the mask, prompt, LTX 2.3 repair LoRA, audio/video latent route, and final reconstruction are connected, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2050245386517921794/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1uMRFBJEPu/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2050245386517921794/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1uMRFBJEPu/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Four-Image Reference Audio-Driven Video Workflow
This workflow is designed for LTX 2.3 four-image reference audio-driven video generation. It combines multiple visual references with audio-aware video latent routing, making it suitable for creators who want a more controlled cinematic video instead of a random text-only result. The main purpose is to use four reference images as visual anchors, then guide the video generation with prompt structure, temporal motion planning, and audio-related conditioning so the final output feels more coherent, more rhythmic, and more production-ready.The workflow is built around the LTX 2.3 video generation system, using LTX video and audio latent components, LTX VAE decoding, Gemma-style text conditioning, custom sampler routes, manual sigma control, and final video export. Compared with a simple image-to-video workflow, this setup is more advanced because it does not depend on only one image. It allows the user to provide multiple reference images that can define different parts of the final video: character identity, product appearance, clothing or pose, background atmosphere, color tone, camera style, and visual direction.The four-image reference structure is the most important visual control layer. In practical use, Image 1 can define the main subject, Image 2 can provide the product or object, Image 3 can guide the scene or environment, and Image 4 can provide the final mood, style, or lighting reference. This gives LTX 2.3 more visual information to work with, reducing the chance of identity drift, unstable product appearance, or inconsistent scene design. For product videos, AI influencer clips, fashion showcases, beauty ads, music-video style shots, and short-form commercial content, this kind of multi-reference structure is much more useful than single-image generation.The audio-driven part makes this version different from the normal four-image reference workflow. The graph includes audio VAE routing, audio latent connection, LTXVConcatAVLatent, LTXVSeparateAVLatent, and LTXVAudioVAEDecode-style processing, allowing the video pipeline to carry audio information through the generation and export process. This makes the workflow suitable for videos where rhythm, performance, presentation timing, music atmosphere, or spoken content matters. It is not just a silent image animation pipeline; it is structured for video output with audio-aware handling.The workflow also includes NAG enhancement, multiple sampling stages, latent upscaling, tiled VAE decoding, and final video creation. These stages help improve visual stability, reduce drifting, refine detail, and make the final output more suitable for publishing. The workflow can generate a first controlled video pass, refine it through later sampling stages, upscale or enhance the latent result, decode the frames, and combine them into a final video output.This workflow is especially useful for AI product advertising, beauty product showcases, character-driven video ads, music-driven AI clips, cinematic image-to-video demonstrations, Civitai workflow previews, RunningHub online demos, YouTube tutorials, and Bilibili content production. If you want to see how the four reference images are connected, how the audio route is handled, and how LTX 2.3 produces the final audio-driven cinematic result, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2052700211776110593/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DERQBeEm1/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2052700211776110593/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DERQBeEm1/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Three-Image Reference Video Workflow
This workflow is designed for LTX 2.3 three-image reference video generation, giving creators a controlled way to turn multiple visual references into a coherent cinematic video. Instead of relying on only one source image, this workflow uses three separate reference images to guide the final result, making it more practical for character consistency, product presentation, scene control, and short-form AI video production.The core idea is multi-reference visual anchoring. A single image often cannot provide enough information for a stable video. It may show the character clearly, but not the product. It may show the lighting, but not the intended camera angle. It may show the scene, but not the subject identity. By using three reference images, this workflow gives the model more visual context. One image can define the main character or subject, the second image can define the product, object, clothing, or key design element, and the third image can provide the background, mood, color palette, or scene atmosphere.This makes the workflow especially useful for commercial-style video generation. For example, creators can use it to build AI influencer clips, beauty product showcases, fashion previews, character-driven advertisements, cinematic product reveals, short social media videos, and Civitai / RunningHub demonstration assets. The prompt can then act as the director, telling the model how the three references should be combined and how the action should develop over time.The workflow is based on an LTX 2.3 video generation route, using image reference guidance, prompt conditioning, video latent creation, sampling, decoding, and final video export. In a typical use case, the reference images are resized and prepared before being passed into the video generation stage. The model then uses these images as guide signals while following the written prompt. This helps the final output stay closer to the intended visual design instead of drifting into a random text-only result.The strength of this workflow is not just reference fusion, but reference fusion for video. In image generation, a reference mismatch may only affect one frame. In video generation, that mismatch can become flicker, identity drift, unstable clothing, object deformation, or inconsistent backgrounds. By giving the workflow three visual anchors, creators can improve the chance that the video keeps a stable subject, more coherent design language, and stronger visual continuity.This workflow is also suitable for PromptRelay-style video planning. The global prompt can define the full scene and visual rules, while local prompt segments can describe the action changes across time. This makes the final video easier to control, especially when the creator wants a product to remain visible, a character to perform a specific action, or the camera to move in a cinematic way.In short, this is a practical LTX 2.3 three-image reference video workflow for creators who want stronger control than single-image video generation, but a simpler setup than larger multi-reference pipelines. If you want to see how the three references are prepared, how the prompt controls the video, and how the final cinematic output is generated, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2052698977556021250/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DERQBeEm1/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2052698977556021250/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DERQBeEm1/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
LTX 2.3 Four-Image Reference Video Workflow
This workflow is designed for LTX 2.3 four-image reference video generation, allowing creators to build a continuous cinematic video from multiple visual references instead of relying on a single image or a text prompt alone. Its main purpose is to use four reference images as visual anchors for scene design, character appearance, product details, atmosphere, and motion continuity, then generate a coherent LTX 2.3 video with stronger control over identity, environment, and visual logic.The workflow uses LTX 2.3 Dev-Dare merged distilled components as the main video generation backbone, together with Gemma 3 text encoding, LTX audio / video latent routing, LTX video VAE decoding, and a custom sampler structure. It also includes enhancement routes such as LTX2.3 Crisp Enhance and VBVR I2V-style LoRA support, making it more suitable for controlled image-to-video generation where visual consistency matters.The key feature is the multi-reference guide structure. Four images can be loaded and processed separately, then resized and prepared through ImageResizeKJv2 and LTXVPreprocess. These references are injected into the generation process through LTXVAddGuideMulti. This gives the model more than one visual source to follow. One image can define the main subject, another can define the product, another can provide the scene or lighting mood, and another can guide the final visual style or atmosphere.The workflow also uses PromptRelayEncodeTimeline, which is important for video control. Instead of writing only one static prompt, the workflow can describe the video as a time-based sequence. The global prompt defines the overall scene, subject, atmosphere, and visual rules, while local prompts describe what happens in different time segments. This makes the final video more cinematic and less random, because the model receives both a stable visual concept and a clear temporal direction.In the uploaded setup, the workflow is structured around a polished beauty product showcase. The prompt describes a young adult East Asian woman presenting a blue skincare serum bottle in a clean studio or cinematic environment, with controlled camera movement, soft light, product visibility, and stable character performance. This shows the intended strength of the workflow: it is not only for abstract animation, but also for commercial-style video generation, product demonstration, AI influencer clips, fashion previews, and short-form advertising content.The workflow includes multiple sampling and refinement stages, manual sigma control, NAG enhancement, latent upscaling, tiled VAE decoding, and final video export. These sections help improve detail, reduce drift, and make the output more usable for publishing. The final result can be exported as a video through the video combine route, making it ready for RunningHub demos, Civitai previews, YouTube showcases, and Bilibili publishing.In short, this is a practical four-image reference LTX 2.3 video workflow for creators who want stronger visual control, better reference fusion, and more stable cinematic video output. If you want to see how the four reference images are connected, how PromptRelay controls the timeline, and how the final video is generated, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2052698989765644290?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1DERQBeEm1/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2052698989765644290?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1DERQBeEm1/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
XiaoYunque Seamless Text Removal | LTX 2.3 Watermark and Subtitle Video Repair Workflow
This workflow is designed for LTX 2.3 video inpainting and authorized video cleanup, focusing on removing unwanted subtitles, watermarks, overlay text, random AI letters, logo marks, and visual text artifacts from a video while keeping the original motion and scene continuity as stable as possible. It is not a simple blur or crop solution. The goal is to reconstruct the damaged area frame by frame so the repaired region blends naturally with the original footage.The workflow uses an LTX 2.3 video repair route with a GGUF-based LTX 2.3 model, LTX23 video VAE, LTX23 audio VAE, LTX conditioning, and a custom sampling structure. It also applies two important LoRA directions: an Edit Anything global LoRA and an inpaint masked R2V LoRA. This combination makes the workflow more suitable for targeted video repair, because the model is guided to understand both the original video context and the masked area that needs to be regenerated.The main logic is mask-based video restoration. The user provides a video and a mask area that covers the unwanted text, subtitle, watermark, or damaged region. The workflow then prepares the video latent, applies the mask as a noise mask, and lets LTX 2.3 regenerate only the target area while preserving the rest of the frame. This is especially important for video repair, because uncontrolled regeneration can easily change the face, background, lighting, camera motion, or scene details outside the repair zone.The prompt and negative prompt are also critical. The negative prompt explicitly suppresses subtitles, captions, Chinese subtitles, watermarks, logos, overlay text, unreadable text, ghost text, flicker, color shift, inconsistent background, blurry patches, and duplicated edges. This helps the workflow understand that the target is clean reconstruction, not adding new text or replacing the scene with unrelated content.The workflow also includes audio-aware routing. Audio can be encoded through the LTX audio VAE and combined with video latent processing, while the final result is exported through VHS_VideoCombine as an MP4 file. This makes the workflow suitable for practical video cleanup instead of isolated frame repair.This setup is useful for repairing your own AI-generated videos, removing accidental prompt text, fixing subtitle contamination, cleaning logo artifacts from authorized material, restoring damaged video areas, and preparing cleaner Civitai, RunningHub, YouTube, or Bilibili showcase clips. If you want to see how the mask is prepared, how the LTX inpaint LoRAs are connected, and how the repaired video is exported, watch the full tutorial from the YouTube link above.⚙️ Try the Workflow Online👉 Workflow: https://www.runninghub.ai/post/2053501073624715266/?inviteCode=rh-v1111Open the link above to run the workflow directly online and view the generation results in real time.If the results meet your expectations, you can also deploy it locally for further customization.🎁 Fan Benefits: Register now to get 1000 points, plus 100 daily login points — enjoy 4090-level performance and 48 GB of powerful compute!📺 Bilibili Updates (Mainland China & Asia-Pacific)If you are in Mainland China or the Asia-Pacific region, you can watch the video below for workflow demos and a detailed creative breakdown.📺 Bilibili Video: https://www.bilibili.com/video/BV1Qa5s6tESM/I will continue updating model resources on Quark Drive:👉 https://pan.quark.cn/s/20c6f6f8d87bThese resources are mainly prepared for local users, making creation and learning more convenient.⚙️ 在线体验工作流👉 工作流： https://www.runninghub.ai/post/2053501073624715266/?inviteCode=rh-v1111打开上方链接即可直接运行该工作流，实时查看生成效果。如果觉得效果理想，你也可以在本地进行自定义部署。🎁 粉丝福利： 注册即送 1000 积分，每日登录 100 积分，畅玩 4090 体验 48 G 超级性能！📺 Bilibili 更新（中国大陆及南亚太地区）如果你在中国大陆或南亚太地区，可以通过下方视频查看该工作流的实测效果与构思讲解。📺 B站视频： https://www.bilibili.com/video/BV1Qa5s6tESM/我会在 夸克网盘 持续更新模型资源：👉 https://pan.quark.cn/s/20c6f6f8d87b这些资源主要面向本地用户，方便进行创作与学习。
Rebels Sulphur 2 GGUF (LTX-2.3 NSFW Model)
GOONERS REJOICELTX-2.3 Sulphur (NSFW MODEL) Distil Workflow — Installation GuideThis workflow runs on smthemex's ComfyUI_LTX2_SM custom node pack with the sulphur_distil distilled transformer GGUF. Below is everything you need to install before queuing the workflow.1. Custom NodesOpen a terminal in your ComfyUI custom_nodes directory and clone:git clone https://github.com/smthemex/ComfyUI_LTX2_SM.git Then install requirements with the embedded Python (portable users):cd ComfyUI_LTX2_SM ..\..\..\python_embeded\python.exe -m pip install -r requirements.txt Or with your venv Python if you're not on portable. Restart ComfyUI fully after install.Make sure ComfyUI itself is updated to the latest stable — older builds won't have the Gemma / GGUF text encoder plumbing this pack relies on.Repo: https://github.com/smthemex/ComfyUI_LTX2_SM2. ModelsAll four core files come from smthem/LTX-2.3-test-gguf: https://huggingface.co/smthem/LTX-2.3-test-gguf/tree/main⚠️ Important: Do NOT use a generic Gemma 3 GGUF from Google, Bartowski, Unsloth, etc. The smthemex loader expects HuggingFace-style tensor names. Standard llama.cpp-format GGUFs will throw an UnboundLocalError: embed_tokens_key on load. Only use the Gemma GGUF from the smthem repo above.Transformer (Sulphur Distil)File: sulphur_distil-Q6_K.gguf (18.1 GB) Folder: ComfyUI/models/gguf/Goes in gguf/, not unet/ or diffusion_models/.OPTIONAL LoRA (FOR VANILLA LTX-2.3)do not use the lora with the model. the lora is for the regular vanilla LTX-2.3 model in case you have too low of vram to run this NSFW model by itself. the lora works, just not as well as the full model does. DO NOT USE THE LORA WITH THE MODEL! lolhttps://huggingface.co/Seregil13th/Sulphur-2-base/blob/main/sulphur_lora_rank_768.safetensorsText Encoder (Gemma 3)File: gemma-3-12b-it-qat-Q4_0.gguf (8.7 GB) Folder: ComfyUI/models/gguf/ConnectorFile: connector.safetensors (6.34 GB) Folder: ComfyUI/models/checkpoints/film_net_fp16.safetensors goes in "frame_interpolation" folderVideo + Audio VAEsTwo options — either source works:Option A — from the smthem repo (same page as everything else):ltx-2.3-22b-distilled_video_vae.safetensors (1.45 GB) → ComfyUI/models/vae/ltx-2.3-22b-distilled_audio_vae.safetensors (365 MB) → ComfyUI/models/vae/Option B — from Kijai's LTX2.3_comfy repo: https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/vaeGrab the matching video and audio VAE files from the vae/ subfolder and drop them in ComfyUI/models/vae/.3. Final Folder StructureComfyUI/ └── models/ ├── gguf/ │ ├── sulphur_distil-Q6_K.gguf │ └── gemma-3-12b-it-qat-Q4_0.gguf ├── checkpoints/ │ └── connector.safetensors └── vae/ ├── ltx-2.3-22b-distilled_video_vae.safetensors └── ltx-2.3-22b-distilled_audio_vae.safetensors 4. Hardware NotesThe smthemex repo lists 6 GB VRAM + 48 GB RAM (peak) as the spec, leaning on the streaming offload code. If you have less system RAM, enable a generous Windows pagefile or you'll hit OOM on the transformer load step. The Q6_K transformer alone is ~18 GB before the encoder and connector come in — there is no shortcut around the RAM requirement.I HIGHLY RECOMMEND UPDATING YOUR BAT FILE WITH THESE FLAGS:--lowvram --disable-xformers --use-pytorch-cross-attention --reserve-vram 2 --disable-smart-memory(these flags will help with the text encoder tricking the push back onto cpu and will burn your vram as priority first)5. TroubleshootingUnboundLocalError: embed_tokens_key → wrong Gemma GGUF. You need gemma-3-12b-it-qat-Q4_0.gguf from the smthem repo specifically. See warning above.Connector not appearing in dropdown → it goes in models/checkpoints/, not models/clip/ or models/text_encoders/.Sulphur GGUF missing from dropdown → it goes in models/gguf/, not models/diffusion_models/ or models/unet/.Out of memory on load → check pagefile size; this workflow benchmarks at ~48 GB peak system memory.If you hit issues outside this list, drop a comment with the full ComfyUI console traceback (not just the popup) and the file you put in each folder.
LTX2.3 Sulphur Multi Frame Video Gen(NSFW)
A workflow uses the latest LTX2.3 Sulphur and Prompt Relay and First and last multi- frames to generate high-quality videos.Other model details are within the workflow.Try the workflow online for free: https://www.runninghub.ai/post/2051993332863254530Please check the video tutorial before you use this workflow: https://rumble.com/v79ibo0-comfyui-ltx2.3-sulphur-video-generation-detailed-guige.html
LTX-2.3, simple workflow. T2V, I2V, I2V external audio, FFLF
Pretty simple workflow for LTXV with 3 modes so far and quite an early version. There is no prompt enchancer; in my opinion, it's useless garbage as a tool, and text LLM work mostly works very slow in ComfyUI. There is no latent upscaling. Almost the same useless garbage.Setting the full resolution will be faster and better than half resolution and latent upscaling with an interpolated result, but it requires resources like second generation. If your system is not powerful enough for the full resolution you set, almost certainly the server will run out of memory on the latent upscale stage. And that's even worse in IMG2video cases with the pixel upscaler itself.The workflow works fine on my 8GB VRAM videocard at 720p; I haven't tried higher resolution yet on FP8 dev model with distilled lora. There is a reason why VAE are separated into loaders and files, I haven't finished that part fully. Controlling it is the same as any other my workflow. If you try even one of my workflows, it will be similar.
No Audio - LTX 2.3 Workflow - Great for B-Roll scenes
Not sure if this is helpful to anyone but,If you want to do video that doesn't need audio: Reaction shots, Fashion Glamour, Nature scenes, I tweaked the Original LTX 2.3 workflow found in ComfyUiRemoved the audio rendering portion(s) of the workflow.(I got tired of prompting "NO AUDIO!" or "NO BACKGROUND MUSIC!" and having LTX not listen.)Added an external LORA loader.Converted the length to seconds instead of framesI found that removing this audio flow speeds video generation up quite well.
LTX Video 2.3 version 1.1 GGUF Image to Video
A GGUF version of the Image to Video workflow of the newest LTX Video version 2.3 in subversion 1.1. HowtoDrag an image into the image nodes, adjust the prompt, and press queue. The rest should fit.DescriptionAn image to video ComfyUI workflow with LTX Video 2.3 subversion 1.1. The nodes are wired and visible in the traditional way and grouped together. No set and get nodes, no subgraph. I like it simple. It produces two videos, one with audio, one without.TimeThe example video in a resolution of 1024 x 576 was generated in 240 seconds at Windows 11 at an AMD Radeon AI PRO R9700. I have yet to test it on Linux, which is usually faster by a third.RequirementsThis workflow was generated with 32 gb vram. There is a ram saver node involved, which you can turn on. So you should be able to go lower. The bottleneck that even made my rig freeze is the VAE decode.
LTX 2.3 Distilled First-Last Frame (GGUF)
VERSION 2Adds prompt relay for better control and adherence, also adds video frame interpolation for smoother frame rates during generation."film_net_fp16.safetensors" goes in:Models/frame_interpolationprompt relay custom node git clone link:git clone https://github.com/kijai/ComfyUI-PromptRelay(Clone to custom_nodes folder)VERSION 1base workflow for ltx 2.3 distilled first last frameFILES:OPTIONAL Kijais fp8 Scaled (requires load diffusion model node instead of unet loader node and replaces the gguf entirely. )https://huggingface.co/Kijai/LTX2.3_comfy/tree/main/diffusion_modelsDistilled ggufshttps://huggingface.co/unsloth/LTX-2.3-GGUF/tree/main/distilledGemma 3_12B FP4 text encoderhttps://huggingface.co/Comfy-Org/ltx-2/blob/main/split_files/text_encoders/gemma_3_12B_it_fp4_mixed.safetensorsAudio VAEhttps://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_audio_vae_bf16.safetensorsVideo VAEhttps://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_video_vae_bf16.safetensorsText Projection text encoderhttps://huggingface.co/Kijai/LTX2.3_comfy/tree/main/text_encodersAbliterated NSFW LoRAhttps://huggingface.co/Comfy-Org/ltx-2/tree/main/split_files/loras
LTX2.3 GGUF Simple i2v + Last Frame 16GB/low vram
diff modelhttps://huggingface.co/unsloth/LTX-2.3-GGUF/blob/main/ltx-2.3-22b-dev-Q4_K_M.gguftext enchttps://huggingface.co/Comfy-Org/ltx-2/blob/main/split_files/text_encoders/gemma_3_12B_it_fp4_mixed.safetensorshttps://huggingface.co/Kijai/LTX2.3_comfy/blob/main/text_encoders/ltx-2.3_text_projection_bf16.safetensorsvaehttps://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_audio_vae_bf16.safetensorshttps://huggingface.co/Kijai/LTX2.3_comfy/blob/main/vae/LTX23_video_vae_bf16.safetensorshttps://github.com/madebyollin/taehv/blob/main/safetensors/taeltx2_3.safetensorslatent upscalerhttps://huggingface.co/Lightricks/LTX-2.3/blob/main/ltx-2.3-spatial-upscaler-x2-1.1.safetensorslora (needed, do not disable)https://huggingface.co/Kijai/LTX2.3_comfy/blob/main/loras/ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105_bf16.safetensors
LtxMTV - 2-Minutes Music Video Generator
LtxMTV is a APP MODE 2-Minutes Music Video Generator(for best results use 50FPS and be patient :) )using Ace1.5 / 4b and Ltx2.3 Distilled FP8both downloadable directly from Comfyui.It's using KJNodes as Custom Nodes.Options:- Auto-Image Resize to 480p or 544p- FPS- BPM- Lyrics- Musical Style- Language- Time SignatureWrite your lyrics, pick your music/singer styleChoose 2 Images, Each scene duration is 20 seconds for a total 6 scenes with alternating between those 2 images as a start frame.Making a good quality music video with minimal effort.for more options, go in Work Flow mode.Tested on a 14900k / nvidia 5080 with 64go Ram
LTX 2.3 - Control Speakers for using Custom Audio
SPEAKER CONTROL with LTX. This workflow might be useful if you have your OWN audio files and you really want to control the speakers by using silence gaps to find speakers.
LTX 2.3 - WW's Distilled Enhancer Experiment
TLDR: A multipass process which potentially improves outputs, particularly from the distilled model.Stemming from my experiments with custom sigma curves, I've stumbled upon an interesting technique which can change the dynamics of an initial video, increasing overall motion, improving fine detail, or adhering to prompts a bit better.Before you get too excited, however, know that it will typically only do one of these things at a time, and sometimes will do the exact opposite. Even when it works, it's only a subtle improvement. This is not the panacea for distilled LTX's shortcomings, but I've found it useful enough that it's become my go-to generator.The way it works is a tad convoluted and is more thoroughly explained by instructions in the workflow itself, but it essentially boils down to the simple expedient of incorporating two extra samplers, powered by the RES4LYF Clown Schedulers, one at low steps and one at low denoising. These help preserve the creativity and obedience of LTX's early steps while preventing too much degradation in overall quality. They are driven by an initial sampling pass at the default values in order to generate usable audio (and an optional template video), and then are further refined by the standard spatial upscaling process.Note that these extra samplers do naturally make the process slower, but since this is all happening with a low step count before the upscaler, it generally only takes about half again as much time as the default workflow.I should also add that this is just an expansion on my general day-to-day workflow and it therefore comes equipped with several functions which can interact with the extra samplers. The greatest impact is on T2V, but I2V, A2V, and V2V are all present in various capacities. I have also included a few switches to partially or completely bypass the extra samplers to make it easier to run comparisons. Or indeed, to use this as a normal workflow, if you happen to like the setup.Which brings us to the clarification that I still have an irrational hatred for subgraphs, so don't necessarily be intimidated by the Great Wall of Nodes, as most of it is just backend. I have included copious notes to try and demystify all adjustable parameters and I advise making judicious use of Comfy's Hide Links function located in the lower right under the minimap.Required custom nodes:ComfyUI-GGUF: https://github.com/city96/ComfyUI-GGUFrgthree-comfy: https://github.com/rgthree/rgthree-comfyComfyUI-KJNodes: https://github.com/kijai/ComfyUI-KJNodesComfyUI-VideoHelperSuite: https://github.com/Kosinkadink/ComfyUI-VideoHelperSuiteRES4LYF: https://github.com/ClownsharkBatwing/RES4LYFAs always, feel free to incorporate or repurpose or whatever you like with this workflow.My thanks again to all those who have uploaded workflows for LTX 2.3. I continue to learn a great deal from you and this certainly wouldn't have been possible without referencing your excellent work.
📜 DaSiWa LTX2.3 Workflows | I2V | FLF2V | T2V | V2V | Audio 📜
My ComfyUI workflows for using LTX 2.3This workflows are used by me to create my art.They are optimized created of my latest knowledge to enhance the outcome. "If this workflow leveled up your day, I'd purr-eciate a like! 😻"Versions & Information👇👇👇👇👇👇👇👇👉 Please read below and the file descriptions "About this version" for more info's.⚠️ Do not use the workflows with the "Nodes 2.0 beta" from ComfyUi or it will mess up things.👇👇👇👇👇👇👇👇What you get from the comfy workflows:♨️ Easy controls✅ As less as possible dependencies🪧 Detailed documentation⛓️ Highly automatic logic✨ Optimized results🔖 Bookmark-Shortcuts with number keysTypes of workflowsOmniForge C-LTX23🎥 I2V, FLF2V, T2V, V2V + Audio🤝 Video resolution matching - Fully automatic scaling⚙️ Chunking (Seconds/Frame based)🎞️ LTX Latent Upscaler (2x native latent)📚 Prompt Enhancer - Automatic prompt refinement🪫 Low VRAM optimizations✨ Multiple Resolution UpscalersTorchlanc (very fast, color correct, low VRAM)Upscale with Model (additional detail, high quality)RTX Super Resolution** (ultra fast, very accurate)🔗 VAE optionsFull VAETiled VAE🔊 Multiple Audio optionsText to Audio (Model audio)Input audio fileVideo audio🫥 Watermark option📢 Soundmark option🧮 Color match feature👾 MiniMeme feature - Create small gif's🃏 Last Frame Extraction🔖 Bookmark-Shortcuts - with number keys🩻 Known issues and advice's⚠️ Some workflows may set on webp av1 encoding (VHS node) - If your computer/setup missing drivers use any other like H265 or H264!Install ffmpeg!Update Comfyui and custom_nodes!Update pytorch 2.9+cu128 or higherMake sure to read where files/models should be placed inside the workflowCheck if the filepath for model/clip/vae match your system like Linux/WindowsSome Custom-Node-Packs need manual installation (e.g. RTX Node)The plugin ComfyUI-DD-Translation can break node connection (avoid)All older Versions are available inside my GitHub Repo.Learned a lot from RuneXX LTX Workflows! They are awesome! Check them out! 🫶Special thanks to @DustyDrab for helping with tests and bug-fixing!YOU are responsible for outputs as always! If you make ToS violating content and I get aware I WILL report this.