OpenRouter Integration

Cync uses OpenRouter as a gateway to access multiple AI models from a single API.

Overview

OpenRouter provides:

• Access to 100+ AI models
• Single API, multiple providers
• Usage-based pricing
• Free tier models available

Setup

1. Get API Key

1. Sign up at OpenRouter
2. Go to API Keys
3. Create a new key
4. Copy the key

2. Configure Environment

# .env.local
OPENROUTER_API_KEY=sk-or-v1-xxx
OPENROUTER_MODEL=anthropic/claude-3.5-sonnet
OPENROUTER_BASE_URL=https://openrouter.ai/api/v1

Model Selection

Free Models (Testing)

Model	Description
meta-llama/llama-3.2-3b-instruct:free	Fast, basic tasks
meta-llama/llama-3.2-1b-instruct:free	Smaller, faster
google/gemma-2-9b-it:free	Google's open model

Paid Models (Production)

Model	Description
anthropic/claude-3.5-sonnet	Best quality
openai/gpt-4o	Excellent reasoning
meta-llama/llama-3.1-70b-instruct	Good balance

Browse all models: https://openrouter.ai/models

API Usage

Chat Completions

const response = await fetch('https://openrouter.ai/api/v1/chat/completions', {
  method: 'POST',
  headers: {
    'Authorization': `Bearer ${OPENROUTER_API_KEY}`,
    'Content-Type': 'application/json',
    'HTTP-Referer': 'https://cync.aicholdings.com',
    'X-Title': 'Cync'
  },
  body: JSON.stringify({
    model: 'anthropic/claude-3.5-sonnet',
    messages: [
      { role: 'system', content: 'You are a helpful assistant.' },
      { role: 'user', content: 'Hello!' }
    ]
  })
});

const data = await response.json();
console.log(data.choices[0].message.content);

With Tool Use

const response = await fetch('https://openrouter.ai/api/v1/chat/completions', {
  method: 'POST',
  headers: { /* ... */ },
  body: JSON.stringify({
    model: 'anthropic/claude-3.5-sonnet',
    messages: [ /* ... */ ],
    tools: [
      {
        type: 'function',
        function: {
          name: 'get_stickies',
          description: 'Retrieve user notes',
          parameters: {
            type: 'object',
            properties: {
              tags: {
                type: 'array',
                items: { type: 'string' },
                description: 'Filter by tags'
              }
            }
          }
        }
      }
    ]
  })
});

Cync Usage

Cync Summaries

Multi-agent pipeline uses OpenRouter for:

• Historian analysis
• Editor composition
• Strategist insights
• Validator scoring
• Polisher refinement

AI Chat

Chat interfaces (/protected/rhea) use OpenRouter for:

• Natural language queries
• Tool-calling decisions
• Response generation

Rate Limits

Rate limits depend on the model and your tier:

• Free models: Generous for testing
• Paid models: Tier-dependent

Check limits: https://openrouter.ai/docs#limits

Handling Rate Limits

if (response.status === 429) {
  const retryAfter = response.headers.get('Retry-After');
  await sleep(parseInt(retryAfter) * 1000);
  // Retry request
}

Costs

OpenRouter uses pay-per-use pricing:

Model	Input (per 1M tokens)	Output (per 1M tokens)
Claude 3.5 Sonnet	$3.00	$15.00
GPT-4o	$5.00	$15.00
Llama 3.1 70B	$0.59	$0.79

Check current prices: https://openrouter.ai/models

Best Practices

1. Use free models for testing - Save costs during development
2. Cache responses - Avoid redundant API calls
3. Monitor usage - Track costs in OpenRouter dashboard
4. Handle errors gracefully - Rate limits, timeouts, etc.