Skip to main content
Base URL: https://platform.runblob.io

POST /v1/veo/generate

Creates a video generation task from text description using Google Veo3.
prompt
string
required
Text description of the video (10-4000 characters)
model_type
string
default:"veo-3-fast"
Model version: "veo-3-fast" or "veo-3-quality"
seed
number
Reproducibility seed (1-10000)
callback_url
string
Webhook URL for completion notifications
bytes_image
string
Base64-encoded image for image-to-video (max 20MB, JPEG/PNG/WEBP)
start_frame
string
Base64-encoded start frame for Frame-to-Frame mode (requires end_frame)
end_frame
string
Base64-encoded end frame for Frame-to-Frame mode (requires start_frame)
reference_images
array
Array of 1-3 base64-encoded images for reference-based generation (only veo-3-fast supported)
filter_text
string
Content filtering options: "moderate", "translate", "moderate_translate"
aspect_ratio
string
default:"16:9"
Video aspect ratio: "9:16" or "16:9".
curl -X POST https://platform.runblob.io/v1/veo/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "A cat playing with a ball in a sunny room",
    "model_type": "veo-3-fast",
    "seed": 1234
  }'

Response

{
  "generation_id": "ffd473d5-5bef-4e14-bf22-8d559be3c19f",
  "status": "pending",
  "seed": 319790131,
  "price": "0.2500"
}
generation_id
string
UUID of the generation task for status checking
status
string
Initial status, always "pending"
seed
number
Random seed used for generation (for reproducibility)
price
string
Amount charged in USD
Video generation typically takes 1-10 minutes. Use the generation_id to check status or set up webhooks for notifications.

Generation Modes

The API supports 4 generation modes - only one mode can be used per request:

1. Text-to-Video

Generate video from text prompt only (no images). Models: veo-3-fast, veo-3-quality

2. Image-to-Video

Animate a single image with text prompt using bytes_image parameter.
  • Formats: JPEG, PNG, WEBP
  • Max size: 20 MB
  • Models: veo-3-fast, veo-3-quality

3. Frame-to-Frame

Generate smooth transition between two frames using start_frame and end_frame parameters.
  • Both frames are required
  • Models: veo-3-fast, veo-3-quality
Frame-to-Frame requires both start_frame and end_frame. Using only one will result in an error.

4. Reference Images

Generate video guided by 1-3 reference images using reference_images parameter.
  • 1-3 images supported (JPEG, PNG, WEBP, max 20 MB each)
  • Only works with veo-3-fast
Validation: Only one mode per request. You cannot combine bytes_image, start_frame/end_frame, and reference_images in the same request.