ComfyUI v0.41.0 adds FLUX.3 Image and MiniMax H3 LightVAE support
ComfyUI v0.41.0 adds native support for Black Forest Labs' FLUX.3 Image models, including the open FLUX 3 Image Klein, and for MiniMax H3's LightVAE. It also adds ImageSDRToHDR and ReferenceMultiLatent nodes.
What it does
This release adds FLUX.3 Image model support to ComfyUI's core, so the open FLUX 3 Image Klein weights run in a local graph, along with MiniMax H3 LightVAE support. New nodes include ImageSDRToHDR, ReferenceMultiLatent and nodes for reading latent geometries.
The bundled templates (v0.11.84) add Klein text-to-image and image-edit workflows for both the base and distilled checkpoints; the distilled one uses 15 steps and no classifier-free guidance.
Results and limits
The release notes list pull requests only and report no measured results. The LoRA stack nodes from v0.40 were briefly disabled in this release and then re-enabled.
Worth knowing
It was published on October 15, 2026 at 22:57 UTC (October 16 at 00:57 Warsaw time). FLUX 3 Image Klein is released under the FLUX Kommunity License, not a fully permissive license.
Try it yourself
Update ComfyUI to v0.41.0, open Templates > Flux.3 Image Klein Distilled: Image Edit, load [your storyboard frame], and prompt: "Change only [element] to [new look]; keep the composition and lighting."