Skip to main content
Available now on ModelsLab · Video Generation

Veo 3.1 API — Google Video GenerationGoogle DeepMind's Veo 3.1 via REST API. Free credits.

Try Veo 3.1API Documentation

Veo 3.1 — three variants, one API surface

Veo 3.1

Google's photorealistic video model

Veo 3.1 is Google DeepMind's latest video generation model — the reference for photorealism and complex scene understanding across long clips.

Veo 3.1 Fast

Quick iteration

Use veo-3.1-fast for quick generations and iterative workflows. Strong quality-to-latency ratio — ideal for production pipelines and A/B testing creative variants.

Veo 3.1 Lite T2V

Cost-efficient text-to-video

veo-3.1-lite-t2v generates 720p-1080p clips from text prompts with lower per-call cost. Great for social video content and long-form scripted output.

Veo 3.1 Lite I2V

Animate stills

veo-3.1-lite-i2v takes an input image and an optional motion prompt. Animate product shots, concept art, or character designs into short cinematic clips.

One REST API

No Google Cloud account

Call Veo 3.1 via the ModelsLab REST API. No Google Cloud signup, no Vertex AI setup, no regional quota management — one ModelsLab key for every video model.

Pricing

Transparent per-clip rates

Veo 3.1 Lite variants on ModelsLab price from ~$0.30 per 5-second clip. Free credits on signup so you can compare against Sora, Kling, Seedance, and Wan before committing.

Webhooks

Async callbacks

Pass a webhook URL and ModelsLab POSTs the completed MP4 URL to your endpoint when generation finishes. Essential for batch video jobs and production pipelines.

SDKs

Python & JavaScript

Official SDKs wrap every Veo endpoint. REST + OpenAPI spec for autogenerated clients in any language.

Endpoints

Multiple endpoints, one model

Veo 3.1 API — Google Video Generation is available through 2 specialized endpoints. Choose the one that fits your workflow.

Veo 3.1 FastVeo-3.1-fastVeo 3.1 Fast Image to Videoveo3.1-fast-image2video

Examples

See what Veo 3.1 can generate

Copy any prompt below and try it yourself in the playground.

Photorealistic landscape

“a golden hour shot of a snow-capped mountain range, dramatic clouds, ultra-photorealistic”

Complex scene

“a bustling street market at dusk, vendors moving, smoke rising from grills, cinematic camera pan”

Product demo

“a luxury perfume bottle rotating on a marble surface, studio lighting, reflections on the glass”

Character scene

“a chef plating a dish in a professional kitchen, hands in focus, steam rising, shallow depth of field”

For Developers

A few lines of code.
Google DeepMind video. Same REST API.

ModelsLab handles the infrastructure: fast inference, auto-scaling, and a developer-friendly API. No GPU management needed.

  • Serverless: scales to zero, scales to millions
  • Pay per clip, no minimums
  • Python and JavaScript SDKs, plus REST API
API Documentation 
import requests


response = requests.post(
    "https://modelslab.com/api/v7/video-fusion/text-to-video",
    json={
  "key": "YOUR_API_KEY",
  "prompt": "a cinematic shot of a snow-capped mountain range at golden hour",
  "model_id": "veo-3.1-lite-t2v",
  "num_frames": 120,
  "output_type": "mp4"
}
)
print(response.json())

FAQ

Common questions about Veo 3.1 API — Google Video Generation

Read the docs

Veo 3.1 is Google DeepMind's latest video generation model, widely regarded as the reference for photorealistic output and complex scene coherence. On ModelsLab you access Veo 3.1 via a single REST endpoint — no Google Cloud or Vertex AI account required.

Veo 3.1 Lite variants on ModelsLab price from ~$0.30 per 5-second 720p clip. Pay-per-call, no monthly commitments, free credits on every new account.

Veo 3.1 improves photorealism on complex scenes, better handling of hands and fine detail, and adds dedicated Lite + Fast variants for different cost/quality tradeoffs. Veo 3 remains available for backward compatibility.

Use veo-3.1-lite-t2v for text-to-video and veo-3.1-lite-i2v for image-to-video on a budget. Use veo-3.1-fast when you need quicker generation times for iterative workflows. All three variants share the same authentication and endpoint shape.

Yes. veo-3.1-lite-i2v animates a still image with an optional motion prompt. Preserves subject identity and composition while adding natural motion.

Each Veo 3.1 call produces 5-10 seconds of video. For longer sequences, chain multiple calls at scene boundaries and stitch the clips together.

Yes. Every new ModelsLab account gets free credits on signup — no credit card required. Enough to test all three Veo 3.1 variants against Sora, Kling, Seedance, and Wan before picking a paid plan.

Yes. Pass a webhook URL in the request body and ModelsLab POSTs the completed MP4 URL to your endpoint when generation finishes — essential for long-running video jobs.

Veo 3.1 leads photorealism. Sora 2 excels at narrative coherence and physics. Kling 3.0 is the reference for cinematic motion realism. All three are available under one ModelsLab API key — route per use case.

Yes. The ModelsLab Python SDK wraps every video endpoint including Veo 3.1. REST + OpenAPI spec for autogenerated clients in any language.

Ready to create?

Start generating with Veo 3.1 API — Google Video Generation on ModelsLab.

Try Veo 3.1API Documentation