Introduction
Generative video gateway for agents and developers — product guides and API reference.
Pipevideo is a generative video gateway. Send an OpenAI Responses-compatible request with a text prompt (and optional reference images or video), and Pipevideo routes the work to an external video orchestration provider. Poll for the result or receive a webhook when generation completes.
Start here
Quickstart
Create an account, issue an API key, and generate your first video.
Core concepts
Organizations, engines, orchestration models, and the generation lifecycle.
Video generation
Create, poll, and webhook flow with parameters and error handling.
API reference
HTTP endpoints, authentication, and the TypeScript SDK.
Who this is for
- Application developers shipping video features without building GPU infrastructure or render pipelines.
- Platform teams routing generation across providers with credit-based billing and usage reporting.
- AI agent builders connecting assistants to video generation via the REST API or MCP server.
What Pipevideo handles
| Concern | Pipevideo |
|---|---|
| Routing | Sends work to an external provider for orchestration and rendering. See Provider routing. |
| Provider selection | Prefer backends with provider.order |
| Billing | Token-based orchestration charges; pay-as-you-go credits |
| Delivery | Poll GET /v1/responses/{id} or receive webhooks |
Browse available engines and per-model pricing on the Engines page.