MiniMax Releases H3 Omnimodal Model: Native Stereo 2K Video Generation
mishig25 · x · 2026-08-30
MiniMax has released MiniMax H3, a general-purpose, omni-modal generative system. The model possesses unified understanding of text, images, video, and audio contexts from the pre-training stage.
Key Specifications:
- Video Output: Native stereo audio, up to 2K resolution, 4-15 seconds duration, 24 FPS.
- Audio: 32 kHz stereo.
- Multimodal Inputs: Supports text-to-video, image-to-video, video-to-video, and combined audio-video generation workflows.
- Languages: Stable support for 11 languages, including Arabic, Chinese, and English.
- Aspect Ratios: Supports a wide range including 21:9, 16:9, 4:3, 1:1, and 9:16.
More from Models
- User feedback: Opus-5 has the lowest token usage of any model used — adonis_singh · 2026-08-30
- heretic: fully automatic censorship removal for LLMs nears 29k stars — p-e-w · 2026-08-30
- Experiment: Claude Easily Assisted in Piracy and Reverse Engineering via agents.md — adonis_singh · 2026-08-30
- OpenAI dominates browser use while Claude's strength is mostly coding, exec says — bindureddy · 2026-08-30
- Model performance degrades in long context; token efficiency varies widely across labs — zakelfassi · 2026-08-30
- Claude Opus 5 Backlash: Benchmarks Soar But Daily Use Fails — gerardsans · 2026-08-30