API Documentation

KiteRouter is an OpenAI-compatible AI gateway for 100+ models. Point an OpenAI-compatible client at KiteRouter's base URL, use a KiteRouter API key, and keep the request shape you already use.

It's Just One Line

diff
1 import OpenAI from 'openai'
2
3 const client = new OpenAI({
4+  baseURL: 'https://kiterouter.com/api/v1',   // ← add this
5-  apiKey: 'sk-...',
6+  apiKey: 'kr-your-api-key', // ← swap your key
7 })

That's it — same SDK and methods, with access to models from OpenAI, Anthropic, Google, and more.

Supported Endpoints

Capability	Endpoint	Notes
Chat completions	`/api/v1/chat/completions`	OpenAI-compatible request and response shape
Responses	`/api/v1/responses`	OpenAI-compatible Responses API
Streaming	Same endpoints with `stream: true`	Server-Sent Events
Models	`/api/v1/models`	Model IDs available to your key

Info: Send OpenAI-compatible request bodies. KiteRouter can route them to compatible upstream providers and map the public model ID to the selected upstream model.

Quick Start

Create an API key in your dashboard
Set the base URL to https://kiterouter.com/api/v1
Use a KiteRouter model ID from Models or /api/v1/models
Call an OpenAI-compatible endpoint with your existing client

bash

1curl https://kiterouter.com/api/v1/chat/completions \
2  -H "Authorization: Bearer kr-your-api-key" \
3  -H "Content-Type: application/json" \
4  -d '{
5    "model": "gpt-4o-mini",
6    "messages": [
7      {"role": "user", "content": "Say hello in one sentence."}
8    ]
9  }'

Base URL

text
1https://kiterouter.com/api/v1

Ready to start building?

Get your API key and make your first request in under a minute.