MiniMax H3 (768p) API

MiniMax's H3 video model at 768p, with sound

minimax-h3-768pAsync taskNative audiovideoLive price on TokPort: $0.030 per second

Calling the API

Base URL https://api.tokport.io, authenticated with Authorization: Bearer <key>. Create a key in your dashboard; the examples read it from TOKPORT_API_KEY. Every response carries an X-Oneapi-Request-Id header you can search in your usage log.

Text to video

POST /v1/videos · async

curl https://api.tokport.io/v1/videos \
  -H "Authorization: Bearer $TOKPORT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"minimax-h3-768p","prompt":"slow dolly-in on a coral reef at sunrise, sunbeams through the water","seconds":"5","ratio":"16:9"}'

# poll until status is "completed", then download
curl https://api.tokport.io/v1/videos/{id} -H "Authorization: Bearer $TOKPORT_API_KEY"
curl https://api.tokport.io/v1/videos/{id}/content -H "Authorization: Bearer $TOKPORT_API_KEY" -o clip.mp4

Asynchronous. The submit response returns a task `id` with status `queued`. Poll `GET /v1/videos/{id}` until `status` is `completed` (progress 0–100), then download bytes from `GET /v1/videos/{id}/content`. Failed tasks are refunded.

Example response

{
  "id": "task_gz9n3dcAaaJ942r9iXjyLzvW7CU9W9eG",
  "object": "video",
  "model": "seedance-2-0",
  "status": "queued",
  "progress": 0,
  "created_at": 1791535189
}

All fields: schema.

Errors

Errors are OpenAI-shaped: {"error": {"message", "type", "code"}}. Failed requests and failed tasks are not billed. See errors & limits.