MiniMax H3, also released as Hailuo 03, turns text, images, video clips, and audio references into 2K video with native stereo sound. Generate 4–15 second clips from a prompt, animate first and last frames, or direct up to fifteen reference files in a single MiniMax H3 request.
1 to 7,000 characters
adaptive works in reference mode only
4 to 15 · default 6
768P or 2K · default 2K · affects price
wire field: content_filter (boolean) · default true · API only
Try one of these prompts
Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.
One sample at the cheapest tier (1 image (6th-9th)).
MiniMax H3 is MiniMax's multimodal video model, also released as Hailuo 03 — both names refer to the same system. On reAPI it runs behind one async endpoint that routes by request shape: a prompt alone for text-to-video, first and last frames for image-to-video, or up to nine reference images, three reference videos and three audio tracks for reference-driven generation. Every request returns 4–15 seconds of video at up to 2K with native stereo audio, billed per second of output plus a small per-image charge for the sixth through ninth reference image.
Real-world workflows and production use cases you can build and ship with this model.
MiniMax H3 builds commercials with varied camera angles, close-up feature shots, readable specifications, and consistent product materials across the whole sequence. Feed product photos in as references and MiniMax H3 keeps packaging, proportions, and brand identity stable from shot to shot.
Try MiniMax H3 now
Direct dialogue, music, ambience, and sound effects in the same prompt as the visuals. MiniMax H3 generates native stereo audio together with the image stream, so lips, footsteps, and score stay synchronized without a separate audio pass.
Assign each asset a role: identity from an image, motion from a video clip, voice from an audio file. MiniMax H3 reference-to-video supports subject replacement, motion transfer, background modification, and relighting — one request, up to fifteen reference files.
Text-to-video, image-to-video, and reference-to-video share one endpoint and one MiniMax H3 model id. The request shape picks the mode — no separate integrations to maintain.
MiniMax H3 renders sound with the picture instead of bolting it on afterwards. Dialogue, ambience, music, and effects follow the same prompt that directs the camera.
MiniMax H3 bills per second of output at one flat 2K rate. No subscription, no idle cost — $1 buys 1,000 credits that never expire, and failed renders refund automatically.
How MiniMax H3 on reAPI compares to Google's Veo 3.1 for prompt-driven video generation with native audio.
Comparison reflects publicly documented behavior at the time of writing. Capabilities and limits change; check each provider's current documentation before choosing.
Create a reAPI account and mint a key in the dashboard. The same key unlocks MiniMax H3 and every other model in the catalog.
OpenOpen the MiniMax H3 playground on this page, pick a mode, and submit a prompt or reference set without writing a line of code.
OpenPOST to /api/v1/videos/generations with model minimax-h3, then poll the task id until the video URL lands.
Opencurl -X POST https://reapi.ai/api/v1/videos/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "minimax-h3",
"prompt": "A cat walking slowly on the beach at sunset, cinematic shot",
"aspect_ratio": "16:9",
"duration": 6
}'Rejected synchronously with the failing field named. Check enum values, ranges, and URL fields — nothing is charged.
401 means a missing or invalid Bearer key; 402 means the reserve exceeds your balance. Manage keys and credits in the console.
Reference material and outputs pass automated moderation. Rejected tasks fail with a clear error and a full refund.
A task that reaches a failed state is never charged — the reserve refunds automatically. Keep the task id and retry when ready.
Common questions about this model.
Try it in the playground or grab an API key to integrate now.