MiniMax H3 Max Turbo

• 🆕 Image-to-video variant in MiniMax's H3 Max Turbo tier, released September 2026
• 🎯 Takes a still image plus a text prompt describing motion
• 🏢 Part of MiniMax's H3 video family, alongside standard and Enhanced tiers
• 🔧 Served through MiniMax's video generation v2 task-creation API
• 💬 Model is selected via a model field when submitting a generation task
• 🌐 Sibling endpoints cover text-to-video and reference-to-video workflows

MiniMax

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…

MiniMax H3 Max Turbo (image-to-video) is the Turbo tier of MiniMax's H3 Max video line, dated September 2026 in this catalog. It is the image-conditioned endpoint: you supply a starting image together with a natural-language prompt describing the motion or action you want. Its same-day sibling MiniMax H3 Max Turbo covers the pure text-to-video path.

The family began with MiniMax H3 and its text-to-video and reference-to-video counterparts, followed by the MiniMax H3 Enhanced revision and then the MiniMax H3 Max tier, with MiniMax H3 Max R2V handling reference-driven generation. This entry sits at the newest point in that progression, adding a Turbo designation on top of the Max tier rather than introducing a new modality.

Access is through MiniMax's video generation v2 API, where a client creates an asynchronous generation task and identifies the desired variant through the request's model field alongside the prompt and input content. Because the H3 line is split by conditioning type, choosing between the image-to-video, text-to-video and reference-to-video endpoints is done at that same field, keeping integration consistent across the family.