AssistantRouter

One API for any LLM

Drop-in OpenAI replacement. Use GPT, Claude, Gemini, or Llama with a single integration. Control costs, switch models, stay compliant.

Get Started FreeRead the Docs
OpenAI-Compatible
Drop-in SDK replacement
Multi-Provider
GPT, Claude, Gemini, Llama
Cost Control
Per-user rate limits
Team Features
EU Mode, priority support
Quick Start
app.ts
import OpenAI from 'openai';

// Use your existing OpenAI code - just change the baseURL
const client = new OpenAI({
  baseURL: 'https://api.assistantrouter.com/v1',
  apiKey: process.env.ASSISTANTROUTER_API_KEY
});

const response = await client.chat.completions.create({
  model: 'gpt-5.2',  // or claude-sonnet-4.5, gemini-3-flash
  messages: [{ role: 'user', content: 'Hello!' }]
});

How It Works

From signup to production in minutes

1

Create workspace

Sign up and create your first workspace. Configure your assistants in the dashboard.

2

Set up assistant

Choose your model, set system prompt, enable tools like RAG and web search.

3

Integrate & ship

Use the OpenAI SDK with our API. Your existing code works out of the box.

Built for Production

Everything you need to ship AI features

OpenAI-Compatible API

Use your existing OpenAI SDK code. Just change the base URL and you're routing through AssistantRouter.

OpenAI
Anthropic
Google
Mistral
Meta
via OpenRouter

Document Search (RAG)

Upload PDFs, DOCX, TXT files. Automatic chunking and retrieval in conversations.

PDFDOCXTXTCSV

Per-User Rate Limits

Set limits on requests, tokens, or cost. Auto-downgrade to cheaper models when exceeded.

Fallback models supported

EU Mode (GDPR)

TEAM+

Route requests through GDPR-compliant providers only. Workspace-level toggle.

GDPR compliant

Web Search

Enable real-time web search for your assistants. Powered by Tavily.

Usage Analytics

Track requests, tokens, and costs. Per-assistant and per-user breakdowns.

Embeddable Widgets

Drop-in chat widgets for your website. Configure domains and styling.

Transparent Pricing

Pay for the platform. Model costs passed through at provider rates.

Hobby
Free
  • 1 assistant
  • 10 RPM
  • 100K tokens/day
  • Community support
View details
Most Popular
Pro
$29/mo
  • 10 assistants
  • 60 RPM
  • 1M tokens/day
  • Web search & RAG
  • Email support
View details
Team
$99/mo
  • Unlimited assistants
  • 300 RPM
  • Unlimited tokens
  • Priority support
  • Team members
View details

Model costs billed separately at provider rates. No markup.

Why AssistantRouter?

Stop building infrastructure. Start shipping products.

Feature
Direct API
AssistantRouter
Multi-provider support
Build each integration
Single API
Model fallbacks
Build from scratch
Built-in config
Per-user rate limits
Manual tracking
Automatic
GDPR routing
Your responsibility
EU Mode toggle
Time to production
Weeks
Minutes

Frequently Asked Questions

You pay a platform fee based on your plan (Hobby is free, Pro is $29/mo). Model costs from providers like OpenAI and Anthropic are passed through at their rates — we don't mark them up. You only pay for what you use.
You can configure fallback models in your assistant settings. If the primary model fails, requests automatically route to your fallback. This is manual configuration, not automatic outage detection.
We're fully GDPR compliant. With EU Mode enabled, requests only go to providers with EU data processing. We don't train on your data or store conversations longer than necessary.
Yes. You can change models at any time through the API or dashboard. Context is preserved between model switches.
We support PDF, DOCX, TXT, CSV, and XLSX files up to 50MB each. Files are automatically chunked and embedded for semantic search.

Ready to ship?

Create your first assistant in minutes. Free to start, scale as you grow.

Get Started FreeRead the Docs