Flux Metal Jacket V15: One 257-Node ComfyUI Workflow Bundling 12 Image Models
CableNo3994 · reddit · 2026-10-12
Flux Metal Jacket V15 ships as a single ComfyUI workflow covering generation, editing, upscaling and finishing, with 12 switchable models: Anima, Boogu, Ernie, Ideogram, Klein 4B/9B, Krea2, Qwen Image, Qwen Image Edit, Qwen Image 2.1, Ming Image and Z-Image Turbo. Its 257 nodes are organized into color-coded groups with per-family toggles so only the needed parts load — handy for comparing and using multiple open-source image models in one place.
More from Multimodal
- Nanobanana 2.1 vs FLUX 3: head-to-head image generation comparison — iamfakhrealam · 2026-10-12
- Sketch-to-3D heads: leaving your sketch unfinished gives H3 more freedom — Suibeam · 2026-10-12
- A short horror film stitched together with Minimax H3, Kroma, VibeVoice, and an LLM agent — abandonedexplorer · 2026-10-12
- Seedance 2.5 demo: elf ranger holding off an army, full prompt included — azed_ai · 2026-10-12
- "AI movies have no soul" is the laziest critique, says filmmaker behind InVideo short 'Fun Size' — churchkey · 2026-10-12
- Endless MiniMax H3 v1.6: ComfyUI workflow for unlimited-length lip-synced video on 12GB — embryo10 · 2026-10-12