UnoRouter: Free Janitor AI Proxy—One Key for All AI Models

Get an OpenAI-compatible key for GPT, Claude, Gemini, DeepSeek, GLM, with a free Janitor AI/SillyTavern proxy. Free models, no credit card.

Visit Website
UnoRouter: Free Janitor AI Proxy—One Key for All AI Models

Introduction

What is it

UnoRouter is a unified AI API platform that gives you access to 300+ AI models through a single API key. It enables you to build apps, run agents, and power various AI workflows—coding assistants, character chats, AI chat clients, and CLI tools—without juggling multiple accounts or SDKs. What makes UnoRouter stand out is its combination of OpenAI-compatible endpoints, provider-agnostic design, and a built-in chat experience that can be used directly on the site. You get the same speed and price across all models, whether you’re using free models (which are included) or paid models billed on a pay-as-you-go basis. Conversations happen in your browser for privacy, and you can bring your own OpenAI-compatible endpoint and key if you prefer. In short, UnoRouter consolidates hundreds of models from dozens of providers under one seamless interface and API.

Features

  • 300+ AI models from 54+ providers accessible via one API key
    • Free models are available and wired in on the backend; paid models are pay-as-you-go
    • Upstreams and provider coverage ensure broad model availability
  • Fully OpenAI-compatible API endpoints
    • Switch providers without changing your code
    • One API key works across coding agents, character chat clients, AI chat clients, and CLI tools
  • Built-in chat experience
    • Open a chat in the UnoRouter UI and talk to ready-made personas or import your own
    • Character chat support with presets, lore books, and import options (e.g., SillyTavern, Janitor.AI, Chub, RisuAI)
    • Conversations stay in the browser; no server-side storage for chats
  • Private & BYOK (Bring Your Own Key)
    • Chats run in your browser; you can connect to your own OpenAI-compatible endpoint and key
    • Security-focused design with optional consent notices for provider inputs
  • No-code chat on the site
    • Free models are ready to chat without signup or key; no setup required
    • Private in-browser chats, with the option to import your own characters or prompts
  • Easy integration and quick start
    • Clear API docs and a minimal curl example to get you started
    • Automatic provider failover and multi-protocol support to maximize uptime
  • Live inference access
    • A wide range of models available for live inferences, including several free options

How to Use

Getting started is a simple three-step process:

  1. Sign up
  • Create an account in under a minute. No credit card is required to start.
  • You can also continue with GitHub, Discord, or Email.
  1. Top up or subscribe
  • Start free with no upfront costs. If you need paid models, top up or choose a plan on-demand.
  • Pay-as-you-go credits never expire, and you can begin using paid models whenever you’re ready.
  • No credit card is required to begin; you only pay when you use paid models.
  1. Get your API key
  • Generate your UnoRouter API key and use it in your code, Claude Code, or paste it into OpenAI-compatible clients like SillyTavern, RisuAI, or any OpenAI-compatible AI chat client.
  • Example API usage (OpenAI-compatible endpoint):
    • POST https://api.unorouter.com/v1/chat/completions
    • Headers: Content-Type: application/json, Authorization: Bearer YOUR_API_KEY
    • Body: {"model": "claude-fable-5.2", "messages": [{"role": "user", "content": "Hello!"}], "stream": true}

Open Chat on the site for immediate interaction:

  • Free models are wired in on the backend, so you can chat right away without signup or key.
  • Pick a character or import your own, and your chats stay in your browser.

Private in-browser chats and BYOK:

  • Conversations are stored in your browser, not on UnoRouter's servers.
  • If you choose a provider that requires input retention or sharing, you’ll see notices and consent prompts.
  • You can also connect to your own OpenAI-compatible endpoint and key and fire requests from the browser.

API-driven workflows:

  • UnoRouter acts as a custom OpenAI-compatible API across coding agents, character chat clients, AI chat clients, and CLI tools.
  • If you’re building agents or chat experiences, you can rely on a single, consistent API surface to reach 300+ models.

Practical use with Janitor AI:

  • Use the provided API key to configure a proxy URL (https://api.unorouter.com/v1/chat/completions) in Janitor AI, select a free model, and refresh the page to start using UnoRouter through Janitor AI.

Pricing

  • Free tier: Access to numerous free models with no cost, subject to per-model usage limits (free models are designed for quick trials and light usage in the chat UI and API).
  • Paid models: Pay-as-you-go pricing. Credits are purchased and do not expire, giving you flexible access to paid models as needed.
  • Startup-friendly: You can begin with zero upfront cost, then top up only when you require paid models. No credit card is required to start.
  • API key usage: After obtaining your API key, you can use it across your codebase, Chat clients, and tools without changing code when switching providers.

Tips

  • Start with free models to prototype and validate your use case before incurring costs.
  • Use provider failover to maximize reliability; if one model is busy, UnoRouter automatically routes to other available models.
  • Take advantage of the in-browser chat to explore character-driven interactions before integrating into apps.
  • When building apps, prefer OpenAI-compatible endpoints to minimize migration friction if you later switch providers.
  • Use BYOK for privacy-sensitive tasks; keep conversations in the browser and only send data to providers when you opt in.
  • Leverage the documentation and quickstart guides to implement your first chat or coding workflow quickly.
  • Monitor model performance and costs by experimenting with free models first, then scaling with paid models as needed.
  • If you’re integrating with a specific tool (e.g., Janitor AI), follow the proxy setup steps to connect quickly and test with a free model.

Frequently Asked Questions

  • Is UnoRouter free? Yes. Many models ending with :free are free to use with usage limits (roughly one request per minute per free model). Paid models are pay-as-you-go and credits never expire. No credit card is needed to start.
  • Is UnoRouter safe to use? Chats in the UnoRouter chat app are stored in your browser, not on UnoRouter’s servers. Some providers may keep or train on inputs; you’ll see notices or consent prompts where applicable. The code is open source.
  • How do I use UnoRouter with Janitor AI? Create an API key, paste https://api.unorouter.com/v1/chat/completions as the Proxy URL in Janitor AI, select a free model, and hard-refresh the page. The Janitor AI Proxy Guide provides field-by-field instructions.
  • Can UnoRouter be down? It’s uncommon, but when free models stop responding, providers for those models may be temporarily busy. Switch to another free model or check the status page for live updates.
  • What can I build with UnoRouter? You can code with coding agents, chat with characters, run AI chats across web UIs, desktop apps, CLI tools, and more—everything from storytelling to software development to data analysis—all through one key.
  • How do I start with the API? Sign up, obtain your API key, and refer to the API docs for Quickstart, base URL, limits, and sample requests. It’s designed to be straightforward to integrate across various clients.
  • Do I need to provide payment information to begin? No. You can start with free models; add credits later for paid models as needed. Credits never expire, and you can top up when you’re ready.

This UnoRouter product overview highlights its core value: a single, unified API to 300+ AI models from 54+ providers, with a built-in chat experience, robust privacy protections, and flexible pricing. Whether you’re a developer building complex agent-based workflows, a product manager exploring multiple AI options, or a creator experimenting with character-driven chats, UnoRouter provides a scalable, provider-agnostic solution that simplifies access to a wide range of models while keeping costs predictable and usage straightforward.