Why Choose EdgeOne Makers as Your Model Router
The smartest llm router for multi-model applications — unified model proxy endpoint, automatic provider routing, secure key hosting, with full SDK compatibility.
One Endpoint, All Models
Access DeepSeek, MiniMax, Hunyuan, OpenAI, Anthropic, Google and more through a single unified API endpoint — switch providers by changing one parameter.
Secure Key Management
Provider API keys are encrypted and hosted by the platform — your application only carries the gateway key, vendor secrets never touch client code.
Zero Platform Fee
The Models gateway service is completely free — no markup, no platform fees. You only pay for vendor API usage on their own platforms when using Bring-Your-Own-Key models.
Native SSE Streaming
All protocols and providers support Server-Sent Events streaming out of the box — real-time responses with OpenAI SDK, Anthropic SDK, or Vercel AI SDK.
How to Use the Model Router in 3 Steps
Connect to the llm router and start multi-model routing immediately — one endpoint, one key, switch providers by changing the model parameter.
1
Get Your API Key
Sign up at the Makers console and create your gateway API key — one key for all model providers.
2
Point Your SDK
Set the base URL to the Makers gateway endpoint in your SDK — supports any OpenAI-compatible client.
3
Call Any Model
Send requests with your chosen `model` parameter — the gateway routes to the right provider automatically.
Complete Model Router Features
Everything your llm router needs for multi-model applications — unified model proxy, provider routing, key management, and streaming support.
Feature | Description |
|---|---|
| Unified Endpoint | Single API endpoint for 8+ LLM providers — switch models by changing one parameter, no code changes |
| Multi-SDK Compatible | Works with OpenAI JS/Python SDK, Anthropic SDK, Vercel AI SDK, cURL, and any OpenAI-compatible client |
| Key Management | Provider keys encrypted and hosted — applications only carry the gateway API key |
| Built-in Models | Call models like `@makers/deepseek-v4-flash` without binding any provider key — free tokens included per account |
| SSE Streaming | Native Server-Sent Events support for all providers — real-time streaming with zero configuration |
What is the model router on EdgeOne Makers?
The model router is a unified llm router that directs API requests to the appropriate LLM provider based on the `model` parameter. It acts as a model proxy between your application and 8+ providers, handling routing, authentication, and response formatting.
How does the llm router handle multi-model switching?
The llm router uses a single endpoint and API key. To switch between providers in a multi-model setup, simply change the `model` parameter in your request — no code changes, no connection reconfiguration needed.
Is the model router free?
Yes. The model router includes 500K free tokens for built-in models. No credit card required. You can also bring your own provider keys to route through the model proxy to any supported provider.
Does the model proxy support all major LLM providers?
Yes. The model proxy routes to DeepSeek, MiniMax, Hunyuan, Zhipu, Moonshot, OpenAI, Anthropic, and Google AI Studio. Multi-model applications can use different providers for different tasks through the same llm router.
Can I use the model router as a drop-in replacement for OpenAI?
Yes. The llm router is fully compatible with OpenAI protocol. Point your OpenAI SDK's base URL to the model router endpoint, and route to any supported provider — making it a seamless multi-model proxy for existing applications.