FluxKleinAllInOne: an all-in-one ComfyUI workflow for generation, reference editing, inpainting and outpainting
755136 · reddit · 2026-10-07
A Reddit user shares their daily-driver ComfyUI workflow FluxKleinAllInOne, combining image generation, reference-based editing, inpainting and outpainting in one place.
Highlights:
- Four modes: reference editing (up to 3 refs), inpainting, outpainting (two modes, one suited for zoom-outs), and free generation
- QwenVL prompt enhancement to auto-describe images into prompts
- Inpaint upscale: inpaint area can be upscaled then composited back for more detail
- Color match: align output colors to a reference or custom image with adjustable strength
Caveats: relies on a dozen+ custom nodes (rgthree, KJNodes, Impact Pack, etc.) and "Show Links" should be off. Full workflow is on Pastebin.
More from Multimodal
- Single-prompt 3D build with Codex, Three.js and fal, prompt shared — OdinLovis · 2026-10-07
- AI-generated robot reggaeton music video 'La Última Ficha' — ScriptLurker · 2026-10-07
- FastH3 V2 generates 5s video with audio in 15s on a single RTX 5090 — NVIDIAAI · 2026-10-07
- FLORA MCP Custom Actions let you define any visual style on demand — round · 2026-10-07
- LINE ZERO: a title sequence built on one visual rule with ComfyUI, LTX and MiniMax H3 — magratheya_64 · 2026-10-07
- SF Tech Week talk: combining LTX and Bria models on the fal platform — OdinLovis · 2026-10-07