$ cat wiki/models/minimax-h3.md
MiniMax H3
Spec
| Attribute | Value |
|---|---|
| Developer | MiniMax |
| Released | 2026-07-31 |
| Announced | 2026-07-31 |
| Context window | unknown |
| Pricing | $0.13 per generated second at 2K · $0.09 per generated second at 768P (768P in closed beta) |
| License | unknown — marketed as open-weight, weights not shipped as of 2026-08-01 |
| Availability | MiniMax platform API (MiniMax-H3), Hailuo AI app |
Context window is unknown: this is a video generation model billed per output | |
| second, and MiniMax's launch material as read states input limits in files (up to | |
| 9 images, 3 videos totalling ≤15s, 3 audio clips; 12 files maximum) rather than in | |
| tokens (source). |
License is unknown rather than a license name because an announced intent to open
weights is not a license — see the end of Release Date.
Release Date
2026-07-31 (source).
MiniMax H3 — widely aliased Hailuo 3.0 / Hailuo 03 after the consumer app it ships in — is a general-purpose omni-modal generation model: one transformer reading text, images, video and audio in a single context and returning video with native stereo sound (source).
Output: native 1440p (2K), 4–15 seconds at whole-second granularity, 24 fps — where Hailuo 2.3 offered a choice between 6 and 10 seconds (source).
Architecture: MiniMax states it set aside the Hailuo-02 architecture. The H3-Omni Transformer separates the understanding and generation workloads during training and tunes hardware utilisation for each, because multimodal context tripled sequence-length variance; end-to-end training throughput rose by nearly 30% (source).
Open weights: announced, not shipped. The model is marketed as open-weight and MiniMax stated an intent to release weights "in the coming days". As of one day after launch, no weights had shipped and the API was the only path (source). This wiki records that as an announced intent rather than a completed release — the same treatment Qwen 3.8 Max (Preview) gets for its "open-weight soon" claim of 2026-07-19, which as of this page's creation has still not been fulfilled. See Open-Weights Policy Fight.
Benchmarks
None published that were readable from this environment, and no third-party measurement exists in this wiki. MiniMax H3 does not appear in the Artificial Analysis leaderboard read 2026-08-02, which covers language models (source). Video generation has no leaderboard in this repo's source set — see Frontier Pacing.
Use Cases
MiniMax positions H3 for advertising, branding, e-commerce, product design, UI/UX and gaming, plus film pre-visualisation and retail catalogue media; named applications include ad variant generation, product and listing videos, animated posters, film title sequences, website hero loops, character-consistent game cinematics and video-to-video motion transfer (source).
The native stereo audio is the operational difference from the previous generation: it is generated alongside the video rather than added in a separate step (source).
Compared To
- Muse Video — Meta's video generation model
- Grok Imagine Video 1.5 (Preview) — xAI's
- Gemini Omni — the other model in this wiki whose pitch is a single transformer across modalities rather than a pipeline
- MiniMax M3 — MiniMax's language/agent model; H3 is a separate line