MiniMax H3 Max
minimax-h3-max-text-to-videominimax/h3-max/text-to-video- 🆕 Max tier of MiniMax's H3 text-to-video line, released August 2026.
- 🏢 From MiniMax, the lab behind the Hailuo video apps.
- 🧠 H3 family is omni-modal: text, images, video, audio in one context.
- 📏 Base H3 outputs up to 15 seconds at up to 2K.
- 🔧 Native stereo audio: dialogue, effects and music generated with the video.
- 🌐 Core H3 weights published under the MiniMax H3 Community License.
- 🎯 Text-prompt entry point; a sibling Max variant covers image-to-video.
- 📚 Max-tier-specific specifications are not detailed in MiniMax's public H3 documentation.
MiniMax is an AI company building generative models across multiple modalities, with a focus that spans both language understanding and audio creation. Their rapid release cadence in early 2026—delivering several new models within just a few months—reflects an ambitious and…
Explore 10 more models by Minimax →MiniMax H3 Max is the top tier of MiniMax's H3 text-to-video family, released in August 2026 alongside its counterpart MiniMax H3 Max (image-to-video). It takes a written prompt and returns a generated clip with synchronized audio, sitting at the head of a line that began with MiniMax H3 in July 2026 and continued with MiniMax H3 Enhanced in early August.
The underlying H3 system is described by MiniMax as a general-purpose, omni-modal generative model: it accepts a unified context mixing text, images, video and audio, and produces video with native stereo audio at resolutions up to 2K and durations up to 15 seconds. Rather than layering a soundtrack on afterwards, dialogue, sound effects and music are modelled jointly with the picture. Reference-driven and first/last-frame modes exist as separate family members, including MiniMax H3 R2V Enhanced.
On MiniMax's own platform, the hosted pipeline adds two server-side stages beyond the raw weights: a context-refinement step that rewrites multimodal instructions into the representation the model consumes best, and a regeneration step that takes a 768p result plus its original context and re-renders it at native 2K. Base H3 weights are openly released under the MiniMax H3 Community License, with regional use restrictions in the license text.
Because MiniMax has not published separate documentation for the Max tier, the quality, speed or duration differences relative to the standard and Enhanced H3 text-to-video variants are best verified directly against the provider's current API reference.
This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.
Daily usage · MiniMax H3 Max · Organic sellers
Top sellers · MiniMax H3 Max · Organic
- Open Forge82.3%
- Venice.ai Proxy17.4%
- attacompute.ai0.2%
- ▲ Apex Ant0.1%
| Seller | Health | Rep↓ | Sales | Video | Routing | $ / s | $ / video |
|---|---|---|---|---|---|---|---|
| Open Forge 0x1d90…b0aa | healthy | 91.11 | $96.65 | 32.2 min | #1 | $0.05 | — |
| Venice.ai Proxy 0x1f22…18c9 | healthy | 86.84 | $20.44 | 16.7 min | #2 | $0.0231 | — |
| ▲ Apex Ant 0x73b4…e736 | — | 81.71 | $0.08 | 5 s | #3 | $0.0162 | — |
| attacompute.ai 0xd5e7…7be0 | — | 51.22 | $0.26 | 5 s | gated | $0.05 | — |
| antseed-tidal-falcon-88cd 0x80d1…88cd | — | 12.78 | — | — | gated | $0.034 | — |
"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.