kenari documentation

kenari is an AI model gateway. You get one API key and one base URL, and you call every chat model in the catalog through OpenAI Chat Completions, Anthropic Messages or OpenAI Responses. Other model types, such as image, speech and music, have their own endpoints in the API reference. You pay in Rupiah, from a prepaid PAYG balance, from a subscription, or with your own provider keys.

Core ideas

• API key: a kn- key from the dashboard. One key works for every model and every API format. See Authentication and API keys.
• Balance and PAYG: you top up a balance in Rupiah, and each request is charged against it by usage. This is called PAYG. See How billing works.
• Model catalog and :free models: Models and pricing lists every model and its price. A model id ending in :free costs nothing to call and has per-minute and daily limits. See Free models.
• Routes: a named chain of models that kenari tries in order, so a failed request moves on to the next step. See Routing and routes.
• Providers: your own API keys and AI subscriptions, such as ChatGPT or Claude credits, used through kenari with no token charges. See Providers.

Choose your path

• Send a first request: follow the Quickstart.
• Use a coding agent: set up Claude Code, Codex or the kenari CLI.
• Build an app: start with Chat completions, then read Function calling, Streaming and Errors.
• Let an AI agent read the docs: see Docs for AI agents and the MCP server.

Base URLs

Client	Base URL
OpenAI SDKs and OpenAI-compatible tools, including Codex	https://kenari.id/v1
Anthropic SDKs and Claude Code	https://kenari.id

Anthropic clients add /v1/messages themselves, so do not add /v1 to their base URL.

Browse the docs

• Get started: begin with the Quickstart.
• Guides: begin with Function calling.
• API reference: begin with Chat completions.
• Account: begin with the Account overview.
• Billing: begin with How billing works.
• Integrations: begin with the Integrations overview.
• Other products: begin with Domains.