API reference

Quickstart

Send your first request with curl, Python, or Node.js.

Base URL

The base URL for every request, the /v1 path rule, and request timeouts.

Authentication

Issuing API keys, the Bearer header, and per-key limits and scopes.

Model selection and access

Model IDs, capabilities, routing, and account or organization access policies.

Endpoints08

/v1/chat/completions

OpenAI-compatible chat completions. Supports streaming, tool calls, and structured outputs.

/v1/messages

Anthropic Messages compatible endpoint and free token counting.

/v1/responses

OpenAI Responses-compatible shim, including background job polling and cancellation.

/v1/images/generations

Image generation. n is at most 10.

/v1/images/edits

Image editing (inpainting and outpainting). Supports both multipart and JSON (base64).

/v1/videos/generations

Video generation (billed per second). Submit, then fetch the result from the dedicated polling endpoint.

GET

/v1/models

List models and retrieve a single model. Slugs containing slashes are supported.

GET

/v1/outputs/{requestId}

Retrieve a stored request and refresh its output URLs, valid for 10 minutes.

Guides05

Streaming

Incremental tokens, SSE completion and errors, and final usage.

Prompt Caching

Cache the repeated prefix to cut input cost — check the cost conditions before you turn it on.

Async Jobs

Responses background jobs, video generation polling, and the webhook alternative in one flow.

Imagen Model Notes

Image count limits and the edit-only model in the Imagen family.

Web Search

The model runs web searches — Google bills per search query; Z.AI bills per search-enabled request.

Integrations04

Claude Code / Anthropic SDK

anthropic-native integration via ANTHROPIC_BASE_URL.

Cursor

Connect by overriding the OpenAI Base URL.

opencode

Example openai-compatible provider configuration.