MiniMax H3-Image open weights: merging 8 classic paintings into one scene
Applecakes_ · reddit · 2026-11-12
After MiniMax released H3-Image's open weights, a Redditor tested its multi-reference capability by feeding in eight classic paintings and asking the model to connect them naturally within one space. The result blends Vermeer's soft realistic lighting with Van Gogh's expressive brushstrokes into a cohesive composition rather than a generic AI blend.
Key points: the model understands relationships across up to 16 reference images instead of copying them into separate sections; generation took 20 seconds per image on an RTX 5060; the ComfyUI BF16 setup uses the pruned H3-Image diffusion model, Qwen3-VL text encoder and FP16 VAE with Euler/Simple sampling, 6 steps, CFG 1; it also supports native 4K generation without upscaling.
More from Multimodal
- Seedance 2.5 Flash goes live on Higgsfield: nearly 2x faster, not yet available in the US — CurieuxExplorer · 2026-11-12
- Blogger chains ChatGPT + Suno + Fable 5.2 to produce a synth-pop music video — technollama · 2026-11-12
- AI-generated breakdance video wows Reddit with fluid motion — TradingCardGirl · 2026-11-12
- One-click RunPod template spins up H3 Editor 0.4.1 with all models — TheTimster666 · 2026-11-12
- Mixed-lighting prompt test: Seedream 5.5 leans ambient, MiniMax H3-Image goes harsh flash — Lopsided-Mud-7769 · 2026-11-12
- Indie studio lands $3,500 client replacing restaurant photo shoots with AI — PrajwalTomar_ · 2026-11-12