Skip to main content
POST
/
v1
/
predictions
curl --request POST \ --url https://api.shuyou.ai/v1/predictions \ --header 'Authorization: Bearer <token>' \ --header 'Content-Type: application/json' \ --data ' { "model": "veo-3.1-generate-preview", "function": "video", "input": { "prompt": "A cinematic shot of a majestic lion in the savannah at golden hour", "aspect_ratio": "16:9", "resolution": "720P", "duration": 8, "generate_audio": true } } '
{
  "data": {
    "task_id": "2c4d50261173430290971a2395a3b607",
    "task_status": "processing"
  }
}

​
Overview

Google Veo 3.1 Generate Preview (veo-3.1-generate-preview) supports:
  • Text-to-video — prompt only
  • First / last frame — first_frame_url, last_frame_url
  • Reference images — up to 3 URLs in reference_image_urls for consistent subject appearance

​
Input parameters

FieldRequiredDefaultNotes
promptYes—Max 1000 characters
aspect_ratioNo16:916:9 or 9:16 only
resolutionNo720P720P, 1080P, 4K
durationNo84, 6, or 8 (seconds)
generate_audioNotrueInclude generated audio
negative_promptNo—Negative guidance text
first_frame_urlNo—First keyframe image URL
last_frame_urlNo—Last keyframe image URL
reference_image_urlsNo[]Up to 3 reference image URLs

Authorizations

​
Authorization
string
header
required

Authorization: Bearer YOUR_API_KEY

Body

application/json
​
model
enum<string>
default:veo-3.1-generate-preview
required

Model ID. Use veo-3.1-generate-preview for this endpoint.

Available options:
veo-3.1-generate-preview
Example:

"veo-3.1-generate-preview"

​
function
enum<string>
required

Task type. Must be video for video generation.

Available options:
video
Example:

"video"

​
input
object
required
​
webhook
string<uri>

Optional HTTPS callback URL when the task completes, fails, or is cancelled.

Response

Async task created

​
data
object
required