macOS Native · 42+ Providers · 100% Local

Smart LLM API Router

A lightweight macOS menu bar proxy that intelligently routes LLM API requests across 42+ providers with automatic failover, protocol conversion, and real-time usage analytics — all running locally on your Mac.

42+
LLM Providers
~8MB
Native Binary
100%
Local & Private
<1ms
Proxy Overhead
Core Features

Everything You Need in One Proxy

SmartLLMRouter replaces multiple tools with a single, lightweight native app.

🔄

Smart Auto-Failover

Priority-based routing with intelligent cooldown — handles 429/5xx/401 errors silently. Three-layer fallback ensures your request never fails.

🔀

Protocol Adapter

Seamless conversion between Anthropic and OpenAI formats. Use Claude Code with OpenAI-compatible providers, or vice versa.

📊

Real-time Analytics

Track daily token usage and estimated costs with 30-day history. Know exactly where your API budget goes.

🔐

Privacy First

100% local execution. API keys stored in macOS Keychain. No telemetry, no cloud sync, no third-party servers.

📦

42+ Providers Built-in

Pre-configured metadata for DeepSeek, OpenAI, Anthropic, Aliyun, MiniMax, and more. One-click setup via providers.json.

Zero Client Config

Just set ANTHROPIC_BASE_URL. Claude Code and other clients work out of the box — no plugins, no wrappers, no config files.

📤

One-Click Config Sharing

Export all channels with optional encryption and share with colleagues. One-click import — no manual re-entry. API keys encrypted, never visible.

Why SmartLLMRouter

Not Just Another Proxy

See how SmartLLMRouter compares to other solutions.

SmartLLMRouterCloud ProxiesOther OSS
Binary Size~8 MB100+ MB50+ MB
macOS Native✓ SwiftUIWeb UIElectron/CLI
Privacy100% LocalCloud RelayedPartial
Auto-Failover3-LayerLimitedManual
Protocol Conversion✓ Built-inPartialNo
Key StorageKeychainServer-sideConfig file
Setup Time< 1 min10+ min5+ min
Use Cases

Built for Real Workflows

01

Claude Code × Multi-Provider

Use multiple API keys for the same model. When DeepSeek hits rate limits, the proxy silently switches to Nvidia or Sensenova — same model, different provider.

export ANTHROPIC_BASE_URL=http://localhost:1897
claude
Claude Code Proxy :1897 DeepSeek Nvidia Sensenova
429 Rate Limited Circuit Breaker Next Provider
02

Rate Limit Failover

When a provider returns 429, the circuit breaker trips and the request automatically retries with the next priority provider. You see no interruption.

03

Context Length Fallback

When a request exceeds a model's context window, SmartLLMRouter intelligently routes to a compatible model with larger capacity — like from gpt-4o (128K) to gpt-4.1 (1M).

200K tokens gpt-4o: 128K ✗ gpt-4.1: 1M ✓
Ollama+ Cloud APIs Unified :1897
04

Local + Cloud Hybrid

Run local models (Ollama, vLLM) alongside cloud APIs. SmartLLMRouter unifies them behind a single endpoint — one URL serves all models.

Architecture

How It Works

Claude Code / OpenAI SDK
SmartLLM Proxy :1897
Protocol Detection
Model Matching
3-Layer Fallback
Protocol Conversion
DeepSeek
OpenAI
Anthropic
Aliyun
MiniMax
Team Workflow

Share Configs, Not Secrets

Export your entire channel setup with one click. Share with teammates — they import in one click. Optional encryption keeps API keys safe.

📤

Export with Confidence

Select channels → Choose encryption (optional) → Save file. All provider configs, model lists, prices, and context lengths are preserved. API keys are encrypted with your password or excluded entirely.

⚙ Channels 🔐 Encrypt 📁 .json
📁 Import File 🔓 Decrypt ✅ Ready
📥

One-Click Import

Drop the exported file → Enter password (if encrypted) → Preview what will be imported → Done. Duplicate detection prevents overwriting existing configs. All channels, models, and metadata are restored instantly.

Ready to Simplify Your LLM Workflow?

Free · Open Source · macOS 13+ · Apple Silicon & Intel