MiniMax H3-Image one-checkpoint 6-step edit demo runs on RTX 3090 in ~60s
Atagor · reddit · 2026-11-13
Reddit user Atagor shares a ComfyUI workflow for one-checkpoint image editing with MiniMax H3-Image: swapping subject, coat, and pose in just 6 sampling steps with scale and grade locked, producing natural results (see attached foggy harbor photo).
- Hardware & quantization: RTX 3090, int8, 60s per image
- Workflow: UNETLoader(int8) → KSampler (euler/simple, 6 steps, cfg 1); CLIPLoader with nvfp4 (type minimax); VAELoader with imagevaefp16 → VAEDecode; two reference images fed via MiniMaxH3ImageConditioning, negative branch handled by ConditioningZeroOut
More from Multimodal
- TikTok creator uses AI to turn marketplace sellers into dogs, topping 15.7M views — gaganghotra_ · 2026-11-13
- LEECKIT V1.4: open-source offline kinetic promo videos from your own images — LeebLaab · 2026-11-13
- Cinematic AI Short Film "Mechanical Metamorphosis" Shared on Reddit — ZidanStudio · 2026-11-13
- "All-in-One Tool for Minimax H3?" User Lists Demanding Video Workflow Wishlist — eeeeekzzz · 2026-11-13
- Ant Group's AntEmbed: Dual-Mode Multimodal Embedding with Fast and Reasoning Modes — _reachsumit · 2026-11-13
- Seedance 2.5 Video Demo Shows Poolside Revenge Scene With Prompt — aftahi_ai · 2026-11-13