Skip to main content
Model logo

MiniMax H3 Highspeed API

  • Text to Video
  • Image to Video
  • Reference to Video

MiniMax H3 Highspeed is part of the MiniMax Hailuo model familyView all MiniMax Hailuo models

Upload images/videos above, then type @ in your prompt to reference them

Tap to uploadPNG, JPG, WEBP, GIF · Max 100MB
Tap to uploadPNG, JPG, WEBP, GIF · Max 100MB
Tap to uploadPNG, JPG, WEBP, GIF · Max 100MB
Tap to uploadMP4, WEBM, MOV · Max 100MB
Output

Your generated video will appear here

Updated

What is MiniMax H3 Highspeed?

MiniMax H3 Highspeed is a faster version of MiniMax H3 that balances quality and speed. It renders 768p video at 24 fps with native audio, runs 4 to 15 seconds per generation, and takes the same inputs as H3: text alone, a start frame with an optional end frame, 1 to 9 reference images, or up to 3 reference videos. It is the model for everyday content where 2K output is not needed and turnaround matters.

Key features of MiniMax H3 Highspeed

Same inputs as H3

Text-to-video, start frame with optional end frame, reference images and reference videos all work the same way as on H3, so a prompt tuned on H3 Highspeed moves to H3 unchanged when you need 2K.

Built for speed at 768p

A faster H3 build with one output resolution, 768p, in six aspect ratios. With a frame image, output follows the input frame shape.

Reference-to-video

Pass 1 to 9 reference images and name them in the prompt as @Image1 through @Image9. The model keeps the subject consistent across generations 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.

Reference videos

Up to 3 reference videos, 15 seconds in total, passed in video_urls. Each second of uploaded video is billed at the per-second output rate.

Native audio in the same pass

Output arrives with an audio track generated alongside the picture, no separate audio step.

What MiniMax H3 Highspeed is best for

  • Everyday content

    Publishable 768p clips with sound without waiting on a 2K render.

  • Batches of social clips

    Fast turnaround on 9:16 and 1:1 clips with native audio.

  • Character-locked series

    Reference images keep the same character across every video in a campaign.

  • Reference-video edits

    Carry the motion or style of a short reference video into a new 768p shot.

MiniMax H3 Highspeed use cases

Produce a week of 9:16 social clips with sound in one batch at 768p. Put a product into a new scene by passing the product and the location as two reference images. Animate a still from a start frame, or pin both ends of the shot with a start and an end frame. Pass a short reference video to carry its motion or style into a new shot.

MiniMax H3 Highspeed limitations

H3 Highspeed outputs 768p only; for 2K use MiniMax H3, and for the cheapest 480p drafts use H3 Lite. It takes no seed, prompt expansion level or reference audio; H3 Fast and H3 Max do. Reference images and frame images are mutually exclusive in one request, and an end frame needs a start frame. Duration is capped at 15 seconds per generation and reference videos at 15 seconds in total.

MiniMax H3 Highspeed API examples

Call MiniMax H3 Highspeed 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-highspeed",
    "input": {
      "prompt": "A hummingbird hovers beside a bright orange hibiscus flower in a sunlit garden, slow push in",
      "aspect_ratio": "16:9",
      "duration": 5
    }
  }'

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.

FAQs

People also ask about MiniMax H3 Highspeed

Speed, resolution and price. H3 Highspeed is a faster version of H3 that outputs 768p only; H3 outputs 768p or 2K. Both take the same inputs: a prompt, a start frame with an optional end frame, up to 9 reference images, or up to 3 reference videos, at 4 to 15 seconds with native audio. H3 also takes a seed, a prompt expansion level and reference audio.

$0.054/sec at 768p, billed per second of output. The first 5 reference images are included and each extra image is billed; a reference video is billed per second of uploaded video at the same per-second rate.

H3 Lite is the budget H3 at 480p or 768p. H3 Highspeed is tuned for faster turnaround at 768p, at a higher per-second rate than H3 Lite. Both take the same text, frame, reference image and reference video inputs.

768p only. aspect_ratio controls the output shape when no start or end frame is supplied; with reference images or videos, auto follows the references.

No. image_urls cannot be combined with start_frame_url or end_frame_url in one request. Use frames for pinned compositions and reference images for subject or style control. An end frame also needs a start frame.

Yes. Audio generates in the same pass as the video, so output arrives with sound and no separate audio step.

Any whole second from 4 to 15 per generation. The default is 5.

MiniMax H3 Highspeed guides and comparisons

Review pricing, limits, and tested outputs before running this model.