mark_k stress-tests new MiniMax H3 Image model with a dense 4K prompt
mark_k · x · 2026-11-12
markk tested the new MiniMax H3 Image model with a deliberately complex 4K test image packed with people, objects, fine details and text. The generated train-station exhibition scene holds up well: signage like "MiniMax H3 Image" is legible, and crowd, robot arms and props stay coherent — a strong sample for a multi-subject + text scene.
More from Multimodal
- Runway open-weights Praxis-1 27B world action model for robots, with NVIDIA — tlakomy · 2026-11-12
- Psychedelic 'Claude Fable 5.2 self portrait' image goes viral on X — repligate · 2026-11-12
- NVIDIA and ElevenLabs expand voice agent partnership with 15MW Blackwell compute and sovereign deployments — nvidia · 2026-11-12
- Seedream V5.5 tested: dozens of characters, hundreds of objects in one complex scene — mark_k · 2026-11-12
- Claude flags mispronunciations in video audio it claims it can't hear — RileyRalmuto · 2026-11-12
- ComfyUI tip: paint inpaint masks as layered PNGs and reuse them across runs — Deftroyy · 2026-11-12