Qwen Image 2.1 Fix v2.0 and custom samplers released for cleaner, less noisy output
Incognit0ErgoSum · reddit · 2026-10-03
A community author released two Qwen Image 2.1 fix models plus a new sampler:
- Qwen Image 2.1 Fix v2.0: surgically removes Qwen's characteristic noise and messy details with almost no effect on composition
- Qwen Image 2.1 Fix Opinionated v1.0: higher quality but alters composition somewhat
The res2mnc sampler (modified from RES4LYF's res2m) and several other Qwen-2.1-tuned samplers are on GitHub (ComfyUI-DPMpp-2M-Sharp, also on the Comfy registry). Weights are downloadable on Civitai and HuggingFace, with the full workflow and prompts included.
More from Multimodal
- Degrading latents don't matter:分段 fresh latents + FL2V 接缝修复跑通长视频 — roychodraws · 2026-10-03
- A meta-prompt that turns any LLM into a YuE2 music prompt builder — Fabulous_Broad_115 · 2026-10-03
- Full prompt shared for a 15-second one-shot magical winter AI video — umesh_ai · 2026-10-03
- Krea 2 Turbo 2-step distillation LoRA hits 4x faster denoising with new checkpoint — TimeTruth2490 · 2026-10-03
- AI-Generated Game of Thrones Anime Trailer Wows With Cinematic Quality — azed_ai · 2026-10-03
- AI video experiment: 'Journey of a morning coffee' camera-movement prompt shared — umesh_ai · 2026-10-03