MiniMax H3 Max Multi-Angle

• - 👁️ Image-to-video camera-control variant of MiniMax's H3 Max line
• - 🆕 Newest H3 Max branch, released September 2026
• - 🧠 Built on H3's omni-modal context: text, images, video, audio
• - 📏 H3 generates up to 2K resolution, 15-second clips
• - 🎯 H3 unifies generation, editing and reference tasks in one model
• - 🔧 H3 released as an open model by MiniMax
• - 💬 Native stereo audio generated alongside video in H3
• - 🏢 From MiniMax, maker of the Hailuo video products

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 Multi-Angle is a specialized image-to-video member of MiniMax's H3 generation, positioned around shot-level camera control. The fal model page for this endpoint describes it as turning a single still image into a clip with keyframe-based control over the camera's orbit, elevation, and distance in 3D space; beyond that, the catalog-level fact is simply that it is the camera-control branch of the H3 Max image-to-video family.

The broader H3 family, announced by MiniMax in 2026, is a general-purpose omni-modal generation model: it jointly understands context spanning text, images, video, and audio, and produces video with native stereo audio at up to 2K resolution and 15 seconds in length, powered by components the company describes as Contextual Omni Representation, H3-VAE, the H3-Omni Transformer, and In-Context Regeneration. MiniMax presents this unified-context design as breaking the boundaries between separate tasks and modalities that previously required distinct models for generation, editing, and reference work.

Within the catalog, Multi-Angle sits alongside the general-purpose MiniMax H3 Max image-to-video endpoint and the latency-oriented MiniMax H3 Max Turbo, and complements MiniMax H3 Max R2V, which conditions on reference material. Compared with the original MiniMax H3 image-to-video release, the Max tier is the later, higher-end branch of the same lineage, and Multi-Angle narrows it further toward directable cinematography.

MiniMax released H3 as an open model, and the vLLM, SGLang, and DiffSynth projects publish their own inference recipes for it. License terms should be checked before deployment.