Ovis-Image

HF | GH
Update your ComfyUI to the latest (Github version) => cd to ComfyUI directory -> terminal -> git pull -> restart ComfyUI

Ovis-Image is this cool 7B text-to-image AI model from Alibaba's team. Basically, it takes your text prompts and whips up images, but it's super good at nailing text inside those images, like making sure words are clear, spelled right, and styled in different fonts without looking messy.

It's awesome for stuff like designing posters, logos, app mockups, or infographics where text needs to pop and be readable, even in long chunks or weird aspect ratios. Handles English and Chinese like a champ, beating bigger models on benchmarks for accuracy and clarity.

Plus, it's efficient, runs on just one beefy GPU with low lag, so great for real-world apps without needing monster hardware. It's solid for general pics too, but text rendering is its killer feature.

These models are redistributed here for the sake of convenience.