Video generation

Veo 3.1

Veo 3.1 is Google's state-of-the-art model for generating high-fidelity, 8-second 720p, 1080p or 4k videos featuring stunning realism and natively generated audio.

Get an API key API examples Videos API docs

From $0.21 per second · 720p, 1080p and 4K · 4, 6 or 8 seconds · Optional audio · MP4 kept 24 hours

PromptA hummingbird hovering at a red flower, macro shot, morning light

Pricing

Price per second

720p · 1280 × 720: $0.21; $0.42 with audio
1080p · 1920 × 1080: $0.21; $0.42 with audio
4K · 3840 × 2160: $0.42; $0.63 with audio

Includes the 5% MiniRouter fee. The final charge uses the provider’s actual cost for the delivered video and stays within your reserved hold. MiniRouter absorbs any overrun. Failed jobs are free.

Estimate your credit hold

Credit is reserved when the job starts. The final charge may be lower.

Size

Duration

With audio

Estimated hold

$1.05

Your length plus 1 second at the per-second price, including the 5% fee. Unused credit is released.

Use the Videos API

Set MINIROUTER_API_KEY to your inference key. Each example starts a job, polls it and saves the MP4.

Language

VIDEO_ID=$(curl --fail-with-body --retry 0 -sS https://api.minirouter.sh/v1/videos \
  -H "Authorization: Bearer $MINIROUTER_API_KEY" \
  -H 'Content-Type: application/json' \
  --data '{
  "model": "google/veo-3.1-generate-001",
  "prompt": "A paper boat drifting down a rainy city street, cinematic.",
  "seconds": 4,
  "size": "1280x720"
}' | jq -r .id)

until curl --fail-with-body -sS "https://api.minirouter.sh/v1/videos/$VIDEO_ID" \
  -H "Authorization: Bearer $MINIROUTER_API_KEY" \
  | jq -e '.status == "completed" or .status == "failed"' > /dev/null; do sleep 10; done

curl --fail-with-body -sS -o video.mp4 "https://api.minirouter.sh/v1/videos/$VIDEO_ID/content" \
  -H "Authorization: Bearer $MINIROUTER_API_KEY"

Retries are off, so a dropped connection cannot start a second job. The MP4 stays available for 24 hours after the job completes.

Operations

Supported operations

Start a job: POST /v1/videos; JSON or multipart fields: model, prompt, seconds, size and optional audio. Prompts up to 2,000 characters.
Check status: GET /v1/videos/{id}; Poll until the status is completed or failed.
Download: GET /v1/videos/{id}/content; Streams the MP4 for 24 hours after the job completes.

Every model, one key.

Get an API key Read the docs