Flux 3 Max

curl --request POST \
  --url https://api.poyo.ai/api/generate/submit \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "model": "flux-3-max/text-to-video",
  "input": {
    "prompt": "A red panda walks through a sunlit forest",
    "duration": 5,
    "resolution": "720p"
  }
}
'

{
  "code": 200,
  "data": {
    "task_id": "task-unified-1792468800-example",
    "status": "not_started",
    "created_time": "2026-10-20T12:00:00"
  }
}

{
  "error": {
    "message": "Invalid request parameters",
    "type": "invalid_request_error"
  }
}

{
  "error": {
    "message": "Invalid request parameters",
    "type": "invalid_request_error"
  }
}

Flux 3 Max creates videos from a text description, animates starting and ending images, or connects images placed on a timeline. Describe the desired scene, motion, camera behavior, and audio in the prompt.

​

Models

• flux-3-max/text-to-video — Video generation from text
• flux-3-max/image-to-video — Video generation from a starting image and an optional ending image
• flux-3-max/keyframes-to-video — Video generation from positioned keyframes

​

Features

• Generate 5–20 seconds of video at 720p or 1080p.
• Generate audio alongside the video or disable it with sound: false.
• Guide transitions with a starting and ending image or up to 10 positioned keyframes.

​

Input modes

Text mode accepts no media fields. Image mode accepts image_urls only, with the starting image first and the optional ending image second. Keyframe mode accepts keyframes only.

​

Required parameters

• prompt (string): Non-empty description of the video, including motion, camera behavior, and desired audio.
• image_urls (array of strings): Required for image mode only. Supply one starting image URL, optionally followed by one ending image URL.
• keyframes (array of objects): Required for keyframe mode only. Supply 1–10 objects containing image_url and frame_index.

​

Optional parameters

• duration (integer): Output duration from 5 through 20 seconds. Defaults to 5.
• resolution (string): 720p (default) or 1080p.
• aspect_ratio (string): auto (default), 21:9, 2:1, 16:9, 4:3, 1:1, 3:4, or 9:16.
• sound (boolean): Generate audio. Defaults to true.

​

Examples

The OpenAPI examples cover text generation, a starting image, starting and ending images, and positioned keyframes. For a transition between two images, select flux-3-max/image-to-video and supply both URLs in image_urls in starting-frame then ending-frame order. For positioned images, select flux-3-max/keyframes-to-video. Set each frame_index on a 24 fps timeline. For a 5-second video, positions can range from 0 through 120, inclusive.

​

Notes

Image URLs must be publicly accessible HTTP(S) URLs for PNG, JPEG, or WebP images. Each keyframe position must be unique and cannot exceed duration × 24. With aspect_ratio: "auto", text mode uses 16:9; image and keyframe modes follow the first image. Keep media fields inside input and callback_url at the request’s top level.