MiniMax H3 (Reference to Video)

POST/v1/run

Generate native-stereo 2K video guided by image, video, and audio references with MiniMax H3.

Request body

Submit an async generation request. The model field selects the model; other fields are model-specific input parameters.

stringmodelrequired

Model identifier. Set to minimax/h3/reference-to-video.

Default: minimax/h3/reference-to-video

stringpromptrequired

Text prompt that describes the image or asset to generate.

Default: Keep the referenced character and motion consistent while creating a cinematic garden scene with synchronized ambient sound.

Optional<array<string>>images

Request parameter supported by this model.

Optional<array<string>>videos

Each clip must be 2–15 seconds; combined duration must not exceed 15 seconds.

Optional<array<string>>audios

Audio requires at least one reference image or video.

integerdurationrequired

Request parameter supported by this model.

Allowed values: 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15

Default: 5

Optional<string>aspect_ratio

Aspect ratio of the generated image.

Allowed values: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16

Default: adaptive

Optional<string>resolution

Output resolution. 768P is generated natively; 2K is upscaled via an in-context regeneration pass.

Allowed values: 768P, 2K

Default: 2K

Response Schema

The submit endpoint returns an accepted generation task. Poll the result endpoint with the returned id for terminal outputs or errors.

Optional<string>error

Structured error returned when the task fails.

stringidrequired

Unique identifier for the submitted request or completed response.

Optional<string>model

Model identifier used for this request.

Optional<array>outputs

Generated output items returned when the task completes.

stringstatusrequired

Current generation status. Non-terminal responses should be polled until completed, failed, or timeout.

Allowed values: pending, running, completed, failed, timeout

Model capabilities

array<string>capability_tagsrequired

Capabilities declared by the model registry.

Default: reference-to-video, video-to-video, audio-to-video

stringexecution_moderequired

Execution mode declared by the model registry.

Default: async

# 1. Submit a generation request
curl -X POST https://api.sandbase.ai/v1/run \
  -H "Authorization: Bearer $SANDBASE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "minimax/h3/reference-to-video",
  "prompt": "Keep the referenced character and motion consistent while creating a cinematic garden scene with synchronized ambient sound.",
  "duration": 5,
  "aspect_ratio": "adaptive"
}'

# 2. If the response is still running, poll the returned id
curl https://api.sandbase.ai/v1/run/run_abc123 \
  -H "Authorization: Bearer $SANDBASE_API_KEY"

Response202 Accepted

{
  "id": "run_abc123",
  "model": "minimax/h3/reference-to-video",
  "status": "running",
  "outputs": []
}