Skip to main content
POST

Create Video (Official Format)

Submit a MiniMax-H3 video generation task, supporting text-to-video, first-frame / first-and-last-frame image-to-video, and reference image/video/audio modes.

Method and Path

Request Example

Response Example

Poll the task_id with Query Task (Official Format) after submission.

Authentication

Body

string
required
Model name. Currently only MiniMax-H3 is supported.
array<object>
required
Multimodal input array; order affects role assignment.
string
required
Content type: text, image_url, video_url, audio_url.
string
Required when type=text. The prompt text.
object
Used when type=image_url; must contain url.
string
required
Public image URL. Formats: JPG, JPEG, PNG, WEBP, HEIC, HEIF; max 30 MB per file; dimensions 256–5760 px; aspect ratio 0.4–2.5. First frame ≤ 1 image, last frame ≤ 1 image, reference images ≤ 9.
object
Used when type=video_url; must contain url.
string
required
Reference video public URL. Formats: MP4, MOV; encoding H.264/AVC, H.265/HEVC; max 50 MB per file; count ≤ 3; each clip 2–15 seconds, total ≤ 15 seconds; dimensions 256–5760 px; aspect ratio 0.4–2.5; frame rate 23.976–60.
object
Used when type=audio_url; must contain url.
string
required
Reference audio public URL. Formats: WAV, MP3; max 15 MB per file; count ≤ 3; each clip 2–15 seconds, total ≤ 15 seconds.
string
Media role:
  • first_frame: first frame (image)
  • last_frame: last frame (image)
  • reference_image: reference image
  • reference_video: reference video
  • reference_audio: reference audio
object
Video generation parameters, all optional.
integer
Video duration (seconds). Valid range [4, 15], default 5.
string
Resolution: 768P, 2K, default 2K.
string
Aspect ratio: adaptive, 21:9, 16:9, 4:3, 1:1, 3:4, 9:16. For text-to-video, defaults to 16:9 and cannot be adaptive; for image-to-video, forced to adaptive.
boolean
Whether to add an AIGC watermark to the generated video, default false.

content Mixing Rules

Violating the following rules may return 400:
  • reference_image / reference_video / reference_audio cannot appear together with first_frame / last_frame
  • Every request must include a non-empty text item (the prompt is required)
  • The total request body must not exceed 64 MB; use public URLs for large files instead of Base64

Generation Mode Mapping

Response

string
Task ID, used for Query Task (Official Format).