Veo 3.1
Veo 3.1 is Google’s latest high-fidelity video generation model that creates short, cinematic clips from text or image prompts with native audio.
What is Veo 3.1?
Veo 3.1 is a state-of-the-art video generation model from Google DeepMind that turns text or image inputs into short, high-quality videos with synchronized audio. It is mainly used for text-to-video and image-to-video generation where creators need precise shot direction, reference imagery, and realistic motion for 4–8 second clips at resolutions up to 4K. It also supports workflows in tools like the Gemini API, Google Vids, and other partner platforms to rapidly prototype ads, social content, and cinematic scenes. Veo 3.1 extends Google’s Veo family of generative video models, succeeding earlier Veo 2 and Veo 3 versions with improved quality, motion, and audio capabilities.
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), or Exacto (highest tool-calling accuracy).
| Provider | Input | Output | Cache read /M | Latency | Throughput | Uptime |
|---|---|---|---|---|---|---|
| ~$2.00/vid | ~$2.00/vid | — | ~2.5s | ~60 vid/min | 100.00% | |
| Vertex AI (Google Cloud) | ~$2.20/vid | ~$2.20/vid | — | ~2.8s | ~45 vid/min | 99.9% |
| Together AI | ~$1.50/vid | ~$1.50/vid | — | ~1.8s | ~80 vid/min | 99.9% |
| Replicate | ~$2.50/vid | ~$2.50/vid | — | ~3.0s | ~40 vid/min | 99.5% |
Try this model
Test Veo 3.1 right here — free to start.
Suggestions for your first prompt
Code snippet
Call the model through the OpenAI-compatible API.
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://inference.example.com/v1"
)
response = client.chat.completions.create(
model="google/veo-3-1",
messages=[
{
"role": "user",
"content": "Describe this image in one sentence."
}
],
)
print(response.to_json())
{
"model": "google/veo-3-1",
"messages": [
{
"role": "user",
"content": "Describe this image in one sentence."
}
]
}
Uptime
Last 30 days
30/30 days operational | 100.00% uptime
5 Core Capabilities
-
Text-to-video
Generates high-fidelity short video clips directly from text prompts, supporting cinematic compositions, varied camera movements, and narrative storytelling control.
-
Image-to-video
Animates one or more reference images into coherent video clips, preserving subjects, style, and composition while adding motion and transitions.
-
Audio + video
Creates videos with native synchronized audio including ambience, sound effects, and dialogue, all guided and timed by the user’s prompt.
-
Scene editing
Edits generated or existing clips with tools like object insertion, extension, and frame-based transitions while maintaining realistic lighting and physics.
-
Vertical storytelling
Produces native 16:9 or 9:16 aspect ratio videos optimized for social platforms, supporting short-form, mobile-first storytelling workflows.
6 Most Valuable Use Cases
- Marketing Video Generation
- Product Demo Videos
- Social Media Clips
- Educational Explainer Videos
- Advertising Creative Production
- Vision-Language Video Research
Why Build on LLM.API?
One unified API. Every major model. Built-in reliability, cost control, and observability.
-
Unified AI Routing
Automatically route each request to the optimal model across providers based on latency, cost, and quality—without changing your integration or redeploying.
One endpoint, all models -
Cost-Aware Orchestration
Dynamically balance premium and budget models using per-call policies and spend limits, so you control performance while keeping infrastructure and experimentation costs predictable.
Cut cost, keep quality -
Resilient Fallback Flows
Define provider-agnostic fallback chains that auto-retry or downgrade across models on timeouts, errors, or quota limits, keeping your product responsive and reliable.
Never drop a request -
End-to-End Observability
Get centralized traces, metrics, and structured logs for every LLM call across providers, with per-model performance, error, and cost breakdowns built in.
See every token -
Task-Level Abstractions
Describe tasks like chat, tools, RAG, or classification once, and let LLM.API standardize prompts and parameters across heterogeneous model APIs.
Code to tasks, not APIs -
High-Throughput Batch Jobs
Run massive batch workloads with built-in queuing, concurrency control, and automatic retries, so you can safely process millions of calls at predictable cost.
Scale to millions of calls
When to Use — When NOT to Use
Use it if...
- You need high-quality text-to-video generation with strong realism and temporal coherence.
- You need to generate short promotional or explainer videos from marketing copy or scripts.
- You need visually rich concept demonstrations, product mockups, or cinematic scenes from prompts.
- Your use case involves iterating on visual storyboards or animatics using natural language edits.
- Your use case involves creative experimentation with camera angles, lighting styles, and visual aesthetics.
- You need AI-assisted ideation for ad creatives, social media clips, or campaign visuals.
Avoid if...
- You need a general-purpose language model for chat, agents, or complex reasoning tasks.
- Your workload requires low-latency, token-level streaming responses for interactive applications.
- You need structured data extraction, code generation, or document understanding rather than video creation.
- Your workload requires on-device or highly resource-constrained inference without powerful GPUs.
- You need strict, fine-grained control over every frame for production-grade animation pipelines.
- You need audio generation, speech recognition, or multimodal conversation instead of video synthesis.
Frequently Asked Questions
-
What is Veo 3.1?
Veo 3.1 is a Google video-generation model accessible via LLM.API, designed to create high-quality, coherent videos from text or image prompts.
-
What is Veo 3.1 best suited for?
Veo 3.1 is best for generating cinematic, longer-form, and stylized videos where temporal consistency and fine-grained visual control are important.
-
What modalities does Veo 3.1 support through LLM.API?
Veo 3.1 supports text-to-video and image-plus-text-to-video generation via LLM.API; it does not handle pure text-chat or audio directly.
-
How is Veo 3.1 priced on LLM.API?
Veo 3.1 pricing on LLM.API is usage-based per video-generation call; check your LLM.API dashboard or pricing docs for the latest unit rates.
-
What is the context window or prompt size for Veo 3.1?
Veo 3.1 accepts relatively long text prompts and optional reference images, but it is not specified in tokens like standard language models.
-
How fast is Veo 3.1, and what latency should I expect?
Veo 3.1 video generations are asynchronous with multi-second to multi-minute latency depending on duration, resolution, and system load.
-
How do I call Veo 3.1 via the LLM.API?
You call Veo 3.1 by selecting the Google Veo 3.1 model in LLM.API, sending a text prompt and optional images to the video-generation endpoint.
-
How does Veo 3.1 compare to other video-generation models on LLM.API?
Compared with many video models, Veo 3.1 emphasizes cinematic quality and temporal coherence, potentially at higher compute cost and latency.
-
What are the main limitations of Veo 3.1?
Veo 3.1 may struggle with precise text rendering, exact physics, copyrighted or unsafe content, and deterministic reproduction of very specific scenes.
-
Can I use Veo 3.1 for real-time or interactive applications?
Veo 3.1 is not suitable for real-time streaming; its generation workflow is batch-oriented with asynchronous result retrieval.
COMPARE
Competitive Models
-
Gemini 3.1 Flash Lite
Gemini 3.1 Flash Lite is Google’s ultra-fast, low-cost Gemini 3-series language model optimized for high-volume, latency-sensitive applications. It prioritizes speed and cost-efficiency while still supporting multimodal understanding and configurable reasoning depth.
-
Gemini 3.5 Flash
Gemini 3.5 Flash is Google’s natively multimodal reasoning model optimized for very low latency and cost while maintaining frontier‑level performance, particularly for coding and agentic workflows.
-
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is a preview large language model from Google’s Gemini family, offering advanced reasoning and multimodal capabilities for early experimentation and feedback. As a preview model, its behavior and performance may change as Google continues development before general availability.
Get one key to every model
Swap your API key. Keep your code.