Open Source Gateway|MIT Licensed|1.6B Free Tokens/Mo

The Intelligent
AI Router

One endpoint for every AI model. Deterministic routing, automatic fallback, and full cost visibility — without vendor lock-in.

Get Started — Open Source (npx routenex)GitHub

cURL Example

OpenAI compatible

# 1. Start gateway instantly (free tier pool ready)
$ npx routenex

# 2. Point any OpenAI SDK or coding agent
$ curl http://localhost:3001/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{"model":"combo/free-unlimited","messages":[{"role":"user","content":"Hello RouteNex!"}]}'

Core Capabilities

Built for Production AI Workloads

Everything you need to orchestrate model providers with reliability, speed, and complete architectural transparency.

Drop-in Replacement

Unified API

One OpenAI-compatible endpoint for OpenAI, Groq, Gemini, Anthropic, Ollama, and more.

Deterministic

Smart Routing

Deterministic routing by cost, latency, or capability. Every decision is explainable and testable.

99.99% Target

Auto Fallback

Circuit breakers and bounded retry chains. When a provider fails, traffic flows to the next healthy target.

Zero Markup

BYOK First

Bring your own API keys. No middleman, no markup, no vendor lock-in. Your keys, your providers.

Built-in Metrics

Full Observability

Request logs, latency tracking, token accounting, and honest cost estimates with provenance timestamps.

Docker Ready

Self-Hosted

Run locally, in Docker, or on your own infrastructure. One command to start. Zero cloud dependencies.

36 Coding Agents & Frameworks Supported

Zero Global Dotfile Pollution

RouteNex launches any terminal coding agent or IDE with scoped tokens and proxy URLs injected strictly per-process. Never overwrite your personal ~/.bashrc or ~/.zshrc.

Claude Code CLI

1-Command

Terminal Agent

routenex run claude --model auto

Cursor IDE

Popular

AI Editor

Override Base: http://localhost:3001/v1

Cline (VS Code)

Native

Autonomous Extension

OpenAI Compatible · Port 3001

Aider CLI

1-Command

Terminal Pair Programmer

routenex run aider --model auto

OpenCode

Python

Code Interpreter

routenex run opencode

Codex CLI

Terminal

CLI Coding Tool

routenex run codex

GitHub Copilot

IDE

IDE Autocomplete

Custom upstream proxy bridge

Continue.dev

Open Source

Open-Source IDE Extension

apiBase: http://localhost:3001/v1

Windsurf

Cascade

Agentic Editor

Custom OpenAI provider

Roo Code

VS Code

Multi-Role Coding Agent

OpenAI Compatible mode

LibreChat

Chat

Self-Hosted AI UI

Reverse proxy /v1 endpoints

Open WebUI

Self-Hosted

Ollama & OpenAI WebUI

OPENAI_API_BASE_URL=...:3001/v1

Dify.ai

Workflow

LLM App Builder

Custom model provider

LangChain

SDK

Agent Framework

ChatOpenAI(base_url="...:3001/v1")

LlamaIndex

SDK

RAG Framework

OpenAI(api_base="...:3001/v1")

LiteLLM

Gateway

Proxy & Gateway

Upstream OpenAI endpoint

CrewAI

Multi-Agent

Multi-Agent Orchestrator

llm = LLM(base_url="...:3001/v1")

AutoGen

Microsoft

Multi-Agent Conversations

config_list: [{"base_url": ...}]

vLLM

Inference

High-Throughput Server

Compatible routing target

Ollama

Local

Local Daemon

http://localhost:11434/v1 bridge

LM Studio

Local

Local GUI

http://localhost:1234/v1 bridge

Jan AI

Offline

Desktop AI

Custom OpenAI server port 3001

LocalAI

Self-Hosted

Drop-in REST API

Upstream gateway adapter

Flowise

Low-Code

Drag & Drop AI Agents

ChatOpenAI Node configuration

AnythingLLM

Desktop

All-in-One Desktop AI

Custom LLM endpoint

Chatbox

Desktop

Desktop Chat App

Custom API host port 3001

NextChat

Web

Cross-Platform Client

Base URL: http://localhost:3001

Bolt.new

Web

In-Browser Full-Stack

Custom backend AI router

Lovable.dev

Web

Full-Stack Agent

Standard API proxy target

v0.dev

Frontend

UI Generation

Unified model upstream

Cursor Composer

IDE

Multi-File Agent

Multi-turn auto-routing

Devin

Agent

Autonomous Software Engineer

Custom base URL config

Kimi Code CLI

OAuth

Moonshot AI CLI

OAuth 2.0 PKCE browser login

Amazon Q

OAuth

AWS Developer Assistant

OAuth 2.0 PKCE browser login

Google Gemini CLI

OAuth

Google Cloud Assistant

OAuth 2.0 PKCE browser login

Antigravity IDE

Native

Advanced Agentic Coding

Built-in auto-discovery bridge

Architecture

How It Works

A lightweight proxy layer sits transparently between your client applications and upstream model providers.

1Client Side

Your Application

Send requests to a single endpoint using any OpenAI SDK or standard HTTP client.

baseURL: http://localhost:3001/v1

2Gateway Engine

RouteNex Router

Validates authentication, checks circuit breakers, executes deterministic routing rules, and manages fallback cascades.

Circuit State:CLOSED (Healthy)

Failover Strategy:Bounded Retry [0ms, 150ms]

3Upstream

AI Providers

Groq, OpenAI, Gemini, Ollama — requests reach the fastest, cheapest, or most capable model with zero code changes.

Groq Llama-3OpenAI GPT-4oGemini 1.5Ollama Local

Primary Path: Sub-5ms Gateway Overhead| Auto-Fallback: Zero Dropped Client Calls

Simple Integration

Start in Seconds

Drop RouteNex into your existing stack by changing only the base URL. Works with any OpenAI-compatible SDK.

curl http://localhost:3001/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{"role": "user", "content": "Explain quantum computing"}]
  }'

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:3001/v1",
    api_key="not-needed"  # optional for localhost
)

response = client.chat.completions.create(
    model="auto",
    messages=[{"role": "user", "content": "Explain quantum computing"}]
)
print(response.choices[0].message.content)

const response = await fetch('http://localhost:3001/v1/chat/completions', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify({
    model: 'auto',
    messages: [{ role: 'user', content: 'Explain quantum computing' }]
  })
});
const data = await response.json();
console.log(data.choices[0].message.content);

Two Product Offerings

RouteNex Product Editions

Deploy RouteNex Open Source on your own servers for free, or leverage RouteNex Cloud for zero-maintenance hosted edge infrastructure.

RouteNex Open Source

Community Edition

$0/forever

Self-host on your own infrastructure or run locally. Full control, zero cloud dependencies, MIT licensed.

All provider integrations (Groq, OpenAI, Gemini, Anthropic, Ollama, Mock)
Deterministic routing engine (cost, latency, balanced)
OpenRouter-style fallback arrays & 3-state circuit breakers
Prompt token compression (10-30% token reduction)
Local Web Admin Console with BYOK key management
Local virtual keys & custom model aliases
Privacy-first embedded telemetry & request inspector
One-command Docker Compose deployment

Get Started — Open Source

RouteNex Cloud

Coming Soon — Waitlist

Managed/private beta

Fully hosted managed AI routing edge with multi-tenant workspaces, team key vaults, and zero server maintenance.

Everything in Open Source
Hosted edge endpoints (https://api.routenex.com)
Multi-tenant team workspaces & RBAC
Centralized organization virtual keys & spend limits
Encrypted cloud credential vaults with automatic sync
Aggregated billing & cost forecasting
Hosted request logs with 90-day retention
99.9% uptime SLA guarantee

RouteNex Cloud (Join Waitlist)

RouteNex Enterprise

Dedicated & VPC

Custom/annual contract

For security-conscious enterprises requiring air-gapped private VPC clusters, custom compliance, and 24/7 SLA.

Everything in RouteNex Cloud
Air-gapped self-hosted deployment options
Dedicated Kubernetes Helm charts & Terraform modules
SAML 2.0 / Okta / Azure AD SSO
SOC2 Type II & HIPAA audit logging
Custom routing plugins & proprietary provider adapters
99.99% uptime SLA with 15-min response guarantee
Dedicated solutions architect

Contact Enterprise Waitlist

The IntelligentAI Router

Built for Production AI Workloads

Unified API

Smart Routing

Auto Fallback

BYOK First

Full Observability

Self-Hosted

Zero Global Dotfile Pollution

Claude Code CLI

Cursor IDE

Cline (VS Code)

Aider CLI

OpenCode

Codex CLI

GitHub Copilot

Continue.dev

Windsurf

Roo Code

LibreChat

Open WebUI

Dify.ai

LangChain

LlamaIndex

LiteLLM

CrewAI

AutoGen

vLLM

Ollama

LM Studio

Jan AI

LocalAI

Flowise

AnythingLLM

Chatbox

NextChat

Bolt.new

Lovable.dev

v0.dev

Cursor Composer

Devin

Kimi Code CLI

Amazon Q

Google Gemini CLI

Antigravity IDE

How It Works

Your Application

RouteNex Router

AI Providers

Start in Seconds

RouteNex Product Editions

RouteNex Open Source

RouteNex Cloud

RouteNex Enterprise

The Intelligent
AI Router