ComfyUI v0.41.3 fixes Flux 3 Klein workflows, adds 9.75GB Qwen3.5-9B prompt upsampler
Mad4reds · reddit · 2026-10-19
Flux 3 Klein launched with a broken ComfyUI workflow; after Black Forest Labs acknowledged it on the Hugging Face repo, ComfyUI shipped v0.41.2 (fixed workflows) and v0.41.3 (updated workflow plus a dedicated prompt upsampler).
Key update details:
- The new recommended workflow requires ComfyUI v0.41.3+ and the flux-3-image-klein-pup prompt upsampler (a Qwen3.5-9B fine-tune; 9.75GB INT8, 18GB full)
- Generating at 2K resolution is recommended to match training conditions
- Model weights are unchanged — only workflows and the upsampler changed
- v0.41.x patch versions were tagged on master without Release entries; v0.42.1 is now the latest release and supersedes them all
- Update via updatecomfyui.bat (Portable), git pull + pip install (Git), or Desktop settings — though Desktop may lag on the stable channel
Related event: BFL and ComfyUI roll out fixes for FLUX 3 Klein launch issues(5 posts)→
More from Multimodal
- PixVerse R2 acts like a world model: WASD movement and text inputs reshape a live game world — hey_abusiddik · 2026-10-19
- FT essay makes the moral case for human music as Suno CEO comments enrage musicians — JMateosGarcia · 2026-10-19
- Dev video: animating Gaussian splats with Spark.js, Three.js and World Labs Marble — theworldlabs · 2026-10-19
- One-shot Star Wars-style 3D scene built by coding agents in ~1.5 hours, almost all code — jasonkneen · 2026-10-19
- Runway launches frontier research design studio betting on real-time generative video — c_valenzuelab · 2026-10-19
- One prompt, snarky Mona Lisa: a quick Argon image-gen demo — fofrAI · 2026-10-19