Skip to main content

​
Quick Introduction

Model Router provides a unified model routing and multimodal task entry point, compatible with mainstream protocols and channels.
  • One entry point covers mainstream models and tasks, compatible with OpenAI-style calls without changing SDKs
  • Preserves native paths for Claude, Gemini, and others to reduce migration and switching costs
  • Built-in stability, quota settlement, notifications, and security policies to help you confidently integrate in production
Base URL: https://app.memorylake.ai — OpenAI-compatible at /v1, Claude-native at /claude/v1, Gemini-native at /gemini/:version. Authenticate with the same sk-… key you use elsewhere on the platform (details).
Looking to add memory to your LLM calls rather than just route them? See Memory Router, which layers persistent memory on top of the same protocols.

​
Quick Start

  1. Create an API Key: Generate an API key in the console
  2. Check Available Models: See which models you can use
  3. Use Your API Key: Choose how to make API calls
  4. Monitor Usage: Track your API usage and billing

​
Documentation

Getting Started

Learn how to create an API key and start using Model Router.

Use API Key

Learn different ways to use your API key: direct requests, code examples, or AI app integrations.

List Available Models

Check which models are available to your API key.

View Usage and Billing

Monitor your API usage and understand billing details.

Error Handling

Learn about common errors and how to troubleshoot them.

Limits and Prerequisites

Understand usage limits and requirements.