LTX-2.5 CQ Enhancer LoRAs: Restore Low-Res Video and Images in ComfyUI
Download the LTX-2.5 CQ enhancer LoRAs: two rank-128 adapters (image + video) that restore and enhance old, low-resolution, poor-quality footage in ComfyUI, with bundled workflows.
CQdesign publishes two rank-128 LoRA adapters for LTX-2.5 that improve, enhance and restore old, low-resolution or poor-quality videos and images. One adapter targets video and the other targets still images. The author describes the result as a generative enhancer rather than a regular upscaler: output is often better than many commercial models and processes faster.
Before / after
Restored example from the model card (left: original, right: enhanced)
Still-image enhancement example
Demo video
Full demo video (also on YouTube)
Weights
| File | Target | Notes |
|---|---|---|
ltx2.5-CQ-enhancer-lora-for-videos-rank128.safetensors | Video | Enhances video clips |
ltx2.5-CQ-enhancer-lora-for-images-rank128.safetensors | Image | Enhances still images |
ComfyUI usage
The LoRAs work with the official LTX-2.5 ComfyUI support. Bundled workflows are in the repo's Workflow/ folder:
Important: use the ltx-2.5-video-vae-conv-bf16.safetensors VAE with these LoRAs (especially for images). The convolution version produces smoother, higher-quality output; the normal VAE gives an over-sharpened look that degrades the image. No prompt is needed for this LoRA.
Availability
- Weights: CQdesign/LTX-2.5-CQ-Video-and-Image-Enhancer-LoRAs
- Demo: YouTube
- Base model: Lightricks/LTX-2.5
Comments
Sign in with GitHub to join the discussion.