MiniMax H3 Image ComfyUI workflow released: 6 steps, up to 20MP, no upscaling
IlovePatata · reddit · 2026-11-14
Reddit user ZETALVX released an optimized ComfyUI workflow for MiniMax H3 Image (on Civitai and GitHub), using the same model and workflow for both generation and editing.
- Setup: RTX 3090 (24GB), minimaxh3imageprunedint8convrot model, qwen3vl32b int8 text encoder, Euler/Simple sampler, just 6 steps, CFG 1, Denoise 1.0.
- Adjustable resolution up to 20MP in the workflow, with no upscaling or prompt enhancement needed.
- A YouTube video demos generation and editing results.
Related event: Optimized ComfyUI Workflow for MiniMax H3 Image Released Open-Source(2 posts)→
More from Multimodal
- Reddit user builds a 6-minute Caesar-at-Alesia animation in Google Flow with consistent characters — EpicFoolss · 2026-11-14
- Open-source Blender add-on runs NVIDIA Kimodo, ARDY and MotionBricks locally for AI motion generation — bdsqlsz · 2026-11-14
- NeurIPS 2025 best paper: diffusion models start memorizing at ~300 steps per image — Mastbubbles · 2026-11-14
- Sci-fi film KILL/SAVE team releases "fake ads from 2045" made with AI video — killsavecom · 2026-11-14
- Stella & Shango Part 2: Viking-themed AI video short continues — stellaXaiXvip · 2026-11-14
- Fable 5.2 trick: one recolored 3D character populates every spectator NPC — chongdashu · 2026-11-14