What is MiniMax Hailuo?

MiniMax Hailuo is MiniMax's video generation lineup, and MiniMax H3, released July 31, 2026, is its current flagship. All four variants share the same /v1/tasks shape: minimax-h3 renders selectable 768p or 2K video at 24 fps with native audio, 4 to 15 second duration, text-to-video, and reference-image control; minimax-2.0 animates from a start frame, an end frame, or both; minimax-2.3 animates a start frame; minimax-2.3-fast runs the same start-frame path with shorter turnaround. Prompts run up to 2000 characters on every variant.

What's new in MiniMax H3

MiniMax H3 supports selectable 768p and 2K output across 16:9, 9:16, 1:1, 4:3, 3:4, and 21:9 aspect ratios. Duration is any whole second from 4 to 15 instead of a 6-or-10 choice. Audio generates in the same pass as the video. And input control widens past frames: 1 to 12 reference images, named in the prompt as @Image1 and up, lock a character, a product or a location across generations. MiniMax's own app also takes reference video and audio for editing; the UnificAlly route does not. On the Artificial Analysis leaderboards (July 31, 2026), MiniMax H3 sits #1 in Video Editing (Elo 1,130), #2 in Text-to-Video with audio at Elo 1,242, behind Google's Gemini Omni Flash, and #3 in Image-to-Video with audio at Elo 1,185.

Key features of MiniMax Hailuo

Reference-to-video on minimax-h3

Pass 1 to 12 reference images and name them in the prompt as @Image1 through @Image12. The model keeps the subject consistent and can combine a character, a product and a location in one shot. Reference images cannot be combined with start/end frames in one request.

Native audio in the same pass

minimax-h3 output arrives with sound generated alongside the picture. No separate audio model, no sync step.

Start and end frame control

minimax-2.0 animates from a start frame, an end frame, or both, at 512p to 1080p. minimax-h3 accepts a start frame, optionally with an end frame.

768p or 2K across six aspect ratios

Choose 768p or 2K at 24 fps on minimax-h3, from 21:9 cinematic to 9:16 vertical. Aspect ratio is selectable for text and reference generation; with a frame image, output follows the input shape.

4 to 15 second duration

Any whole second from 4 to 15 on minimax-h3. The 2.x variants keep the 6-or-10 choice, with 1080p locked to 6 seconds.

What MiniMax Hailuo is best for

Character-locked series
minimax-h3 with subject reference images keeps the same character across every video in a campaign.
Character-in-location shots
minimax-h3 with the character as @Image1 and the setting as @Image2 places one inside the other.
Cinematic 21:9 spots
The 2944x1248 widescreen tier renders trailer-shaped output straight from a text prompt.
Start-and-end frame ads
minimax-2.0 or minimax-h3 with matched hero frames when you already have the open and close as key art.
Sound-on social video
Native audio on minimax-h3 means output is publishable without an audio pass.
Quick image-led shoots
minimax-2.3-fast when you have a plate and need motion quickly on a tight latency budget.

MiniMax Hailuo variants

Hailuo has four models on the same API, and each has its own page with full details, playground, and FAQs.

minimax-h3

The flagship. Selectable 768p or 2K output at 24 fps, native audio, 4 to 15 second duration, six aspect ratios, and three input styles: text alone, a start frame with an optional end frame, or 1 to 12 reference images. The prompt is required on this model.

minimax-2.0

The start-and-end frame model. Pass a starting frame, an ending frame, or both with a prompt, and the model animates the connective motion. At least one frame is required. Resolution options are 512p, 768p, and 1080p.

minimax-2.3

The start-frame image-to-video model. Pass a start frame and a prompt describing the motion. There is no text-only mode and no end frame.

minimax-2.3-fast

The faster start-frame model. It takes the same start frame and prompt as 2.3, and turnaround is shorter. Use it when the workflow is image-led and the latency budget is tight.

MiniMax Hailuo use cases

Run a product campaign where the same mascot appears in ten videos by sending the character sheet as reference images to minimax-h3. Put a product into a new scene by passing the product and the location as two reference images. Render a 21:9 teaser at 2944x1248 from a text prompt for a landing-page hero. Build a 6-second 1080p ad with a fixed open and close through minimax-2.0 with start and end frames. Animate stills on minimax-2.3, and route high-volume image-to-video jobs through minimax-2.3-fast.

MiniMax Hailuo limitations

MiniMax H3 caps at 15 seconds per generation and 24 fps, with 768p and 2K output but no 4K tier; rivals like Kling 3.0 render native 4K. Reference images and frame images are mutually exclusive in one request, and the UnificAlly route takes no reference videos or audio. Only minimax-h3 runs from a prompt alone. On the 2.x variants, 1080p locks duration to 6 seconds, and 10-second runs drop to 768p.

MiniMax Hailuo API examples

Call MiniMax Hailuo from any language by POSTing to /v1/tasks. Full parameter docs live at docs.unifically.com/models/video/minimax/hailuo.

curl -X POST https://api.unifically.com/v1/tasks \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "hailuo/minimax-h3",
    "input": {
      "prompt": "A slow dolly shot across a rain-soaked neon street, reflections rippling in the puddles",
      "aspect_ratio": "21:9",
      "resolution": "2K",
      "duration": 10
    }
  }'

Successful submission returns a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the finished result.

Preguntas frecuentes

Preguntas frecuentes sobre MiniMax Hailuo

MiniMax H3, released July 31, 2026, adds selectable 768p and 2K output at 24 fps, stretches duration to any length from 4 to 15 seconds, adds 21:9 and three more aspect ratios, generates native audio in the same pass, and adds omni-reference generation with up to 12 images.

Default to minimax-h3 for quality, 2K output, native audio, reference images, and prompt-only text-to-video, which only H3 offers. Use minimax-2.0 when you need an end frame on a budget. Use minimax-2.3 for start-frame image-to-video, and minimax-2.3-fast for the same input with a tight latency budget.

minimax-h3 renders selectable 768p or 2K output at 24 fps for 4 to 15 seconds. The resolution parameter selects the output tier, while aspect_ratio selects the shape when no start or end frame is supplied. The 2.x variants render 768p or 1080p at 6 or 10 seconds; 1080p only supports 6 seconds, and minimax-2.0 adds 512p.

MiniMax H3 does. Audio generates in the same pass as the video, so output arrives with sound and no separate audio step. The 2.x variants output silent video.

On the Artificial Analysis boards (July 31, 2026), MiniMax H3 is #1 in Video Editing at Elo 1,130, #2 in Text-to-Video with audio at Elo 1,242, and #3 in Image-to-Video with audio at Elo 1,185. Video editing is not exposed on the UnificAlly route.

Yes, on all four variants (1 to 2000 characters). minimax-h3 can run on the prompt alone. minimax-2.0 also needs a start frame, an end frame, or both. minimax-2.3 and minimax-2.3-fast also need a start frame.

Up to 2000 characters across all four variants.

There is no free tier. New UnificAlly accounts get a $0.50 signup balance, enough for about 29 seconds of video on Hailuo 2.0 at 512p ($0.0169 per second). Unlike daily credit allowances elsewhere, the balance never expires. After it runs out, billing is pay-per-use with no subscription.