Introduction

The Pipevideo API is a REST interface for video generation. It supports two OpenAI-compatible request shapes:

Responses API — input + optional instructions (recommended for new integrations)
Chat Completions API — messages array (for existing chat.completions.create clients)

Both create the same async video generation and return a generation id to poll or receive via webhook.

Base URL: https://api.pipevideo.co/v1

Base URL, content types, and request conventions.

API keys and Bearer token authentication.

Create and retrieve video generations (recommended).

OpenAI chat completions shape for existing integrations.

npm install pipevideo — typed client for Node.js and edge runtimes.

Endpoints

Method	Path	Auth	Description
`GET`	`/v1/engines`	Public	List engines and orchestration models
`POST`	`/v1/responses`	API key	Create a video generation (Responses API)
`GET`	`/v1/responses/{id}`	API key	Get generation status and result
`DELETE`	`/v1/responses/{id}`	API key	Cancel an in-flight generation
`POST`	`/v1/chat/completions`	API key	Create a video generation (Chat Completions API)
`GET`	`/v1/chat/completions/{id}`	API key	Get generation status and result
`DELETE`	`/v1/chat/completions/{id}`	API key	Cancel an in-flight generation
`GET`	`/v1/usage`	API key	Usage stats, balance, and breakdown

Approach	Best for
REST API	Any language, full control
TypeScript SDK	Node.js and edge runtimes with typed helpers
OpenAI SDK	Existing OpenAI integrations — use `responses.create` or `chat.completions.create`
MCP server	AI assistants and agent frameworks