Base URLhttps://api.unli.dev/v1

Overview

UNLI serves unlimited tokens through OpenAI- and Anthropic-compatible endpoints. Ask for the auto model and each request is routed to a suitable model. The rest of your code stays the same.

chat.ts

 import OpenAI from "openai";
 
 const client = new OpenAI({
−Removed:   baseURL: "https://api.openai.com/v1",
+Added:   baseURL: "https://api.unli.dev/v1",
−Removed:   apiKey: process.env.OPENAI_API_KEY,
+Added:   apiKey: process.env.UNLI_API_KEY,
 });
 
 const completion = await client.chat.completions.create({
−Removed:   model: "gpt-4o",
+Added:   model: "auto",
   messages: [{ role: "user", content: "Hello, AI!" }],
 });

Get an API key Read the quickstart

Endpoints

UNLI exposes two conversation endpoints, one per API format, plus a model listing.

POST/v1/chat/completionsChat Completions · OpenAI format POST/v1/messagesMessages · Anthropic format GET/v1/modelsModels · Available models and their capabilities

What you get

Unlimited tokens: Build without tracking usage limits.
Automatic model routing: Set model to auto (or unli-auto on Messages) and UNLI picks a model for each request.
Drop-in compatibility: Works with the official OpenAI and Anthropic SDKs. Change the base URL and API key.
Streaming: Real-time responses over Server-Sent Events (SSE).
Vision: Analyze and describe images sent by URL or base64.
Tool calling: Function calling and structured outputs.
Multi-turn context: Conversation context carries across turns.
Sampling controls: Tune output with temperature, top_p and top_k.
Low latency: Optimized for fast response times.

Overview

Endpoints

What you get

On this page