Platform / API Gateway

One API. Every LLM.

A unified gateway to all LLM providers. Smart routing, automatic failover, and built-in optimizations—without changing your code.

Simplified architecture

Replace complex integrations with a single, intelligent endpoint.

Your Application

Single API endpoint

Unified response format

Metriqual Gateway

• Intelligent routing

• Automatic failover

• Cost optimization

• Quality monitoring

O
A
G
M
C
M

30+ Providers

All major LLMs

Custom endpoints

Enterprise-grade gateway features

Smart Routing

Automatically route requests to the best model based on cost, quality, and latency requirements.

Automatic Retries

Built-in retry logic with exponential backoff. Never lose a request to transient errors.

Load Balancing

Distribute load across multiple providers. Avoid rate limits and maximize throughput.

Fallback Models

Define fallback chains. If GPT-4 fails, automatically try Claude or another model.

Rate Limiting

Control costs with built-in rate limiting. Set limits per user, API key, or globally.

Global Edge Network

Route requests through the nearest edge location for minimal latency worldwide.

Provider status

Real-time status of all supported LLM providers

OpenAI

GPT-4, GPT-3.5, DALL-E

operational

Anthropic

Claude 3 Opus, Claude 3 Sonnet

operational

Google

Gemini Pro, PaLM 2

operational

Mistral

Mixtral, Mistral 7B

operational

Cohere

Command, Embed

operational

Meta

Llama 2, CodeLlama

operational

Simple integration

Use any LLM provider with the same clean API

unified-api.js
// One endpoint for all providers
const response = await fetch('https://api.metriqual.com/v1/chat/completions', {
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    // Just change the model to switch providers
    model: 'gpt-4',           // OpenAI
    // model: 'claude-3',     // Anthropic
    // model: 'gemini-pro',   // Google
    // model: 'mixtral',      // Mistral
    
    messages: [{
      role: 'user',
      content: 'Hello!'
    }],
    
    // Optional: Let us pick the best model
    routing: {
      mode: 'auto',
      optimize_for: 'quality',  // or 'cost' or 'speed'
      max_cost: 0.10
    }
  })
})

// Consistent response format across all providers
const data = await response.json()
console.log(data.choices[0].message.content)

Simplify your LLM infrastructure

Stop managing multiple SDKs and endpoints. Start building with a unified API.