openrouter

OpenRouter API Example

A multi-agent example using OpenRouter API as the LLM provider, demonstrating how to integrate with different LLM services and leverage OpenRouter's model selection capabilities.

Tech Stack and Components

Technologies Used

• Language: Go 1.26+
• LLM Provider: OpenRouter API (unified access to multiple models)
• Configuration Format: YAML
• Concurrency Control: errgroup
• Template Engine: Go text/template
• HTTP Client: OpenRouter API calls

Core Components Used

OpenRouter Features

Key Feature Implementations

Code Location References:

• OpenRouter configuration loading: examples/openrouter/main.go:30-50
• Leader Agent creation: examples/openrouter/main.go:100-120
• Sub Agents creation: examples/openrouter/main.go:125-145
• API calls: internal/llm/client.go:80-120
• Error handling: internal/llm/client.go:150-180

OpenRouter API Integration

Get API Key

1. Visit OpenRouter
2. Register account and log in
3. Get your API key

Supported Models

OpenRouter supports multiple models, commonly used ones include:

Quick Start

1. Set API Key

export OPENROUTER_API_KEY="sk-or-v1-your-api-key"

Or set in config/server.yaml:

llm:
  api_key: "sk-or-v1-your-api-key"

2. Configure Model

Edit config/server.yaml:

llm:
  provider: "openrouter"
  api_key: "${OPENROUTER_API_KEY}"  # Use environment variable
  base_url: "https://openrouter.ai/api/v1"
  model: "meta-llama/llama-3.1-8b-instruct"
  timeout: 60
  max_tokens: 2048

3. Run Example

cd examples/openrouter
go run main.go

4. Example Output

=== Style Agent (OpenRouter Example) ===
Starting Style Agent Example

[INFO] Loaded configuration from ./examples/openrouter/config/server.yaml
[INFO] Using OpenRouter API with model: meta-llama/llama-3.1-8b-instruct
[INFO] Initialized Leader Agent: leader-openrouter
[INFO] Initialized 3 Sub Agents

=== Processing Sample Request ===
Input: I want to travel to Tokyo, Japan for 5 days and 4 nights, with a budget of 10,000 yuan

[Leader Agent] Parsing profile using OpenRouter...
Profile: {"destination": "Tokyo", "duration": "5 days and 4 nights", "budget": 10000}

[Leader Agent] Dispatching tasks to 3 sub-agents...

[Agent: destination] Processing with OpenRouter...
[Agent: food] Processing with OpenRouter...
[Agent: hotel] Processing with OpenRouter...

=== Aggregated Results ===
Destination Recommendations: ...
Food Recommendations: ...
Hotel Recommendations: ...

[INFO] Example completed successfully

Configuration

config/server.yaml

llm:
  provider: "openrouter"
  api_key: "${OPENROUTER_API_KEY}"  # Recommended: use environment variable
  base_url: "https://openrouter.ai/api/v1"
  model: "meta-llama/llama-3.1-8b-instruct"
  timeout: 60
  max_tokens: 2048

agents:
  leader:
    id: "leader-openrouter"
    max_steps: 5
    max_parallel_tasks: 3

  sub:
    - id: "agent-destination"
      type: "destination"
      triggers: ["destination", "go", "travel"]
      max_retries: 2

prompts:
  profile_extraction: |
    Extract travel information from user input: {{.input}}

OpenRouter Features

1. Model Selection

You can choose different models in configuration:

llm:
  model: "meta-llama/llama-3.1-70b-instruct"  # More powerful model

2. Cost Tracking

OpenRouter provides detailed cost tracking, you can view costs for each request:

# View usage in OpenRouter console
# https://openrouter.ai/activity

3. Error Handling

The code includes comprehensive error handling:

// internal/llm/client.go:150-180
if err != nil {
    if strings.Contains(err.Error(), "401") {
        return fmt.Errorf("API key invalid")
    }
    if strings.Contains(err.Error(), "429") {
        return fmt.Errorf("rate limit exceeded")
    }
    return fmt.Errorf("API error: %w", err)
}

Differences from Other Examples

Troubleshooting

Issue 1: Invalid API Key

Error: API key invalid

Solution:

• Check OPENROUTER_API_KEY environment variable
• Verify API key is correct
• Confirm API key is activated

Issue 2: Model Not Available

Error: model not found or not available

Solution:

• Check model name is correct
• Visit OpenRouter Models to view available models
• Try using a different model

Issue 3: Rate Limit

Error: rate limit exceeded

Solution:

• Wait and retry after some time
• Upgrade OpenRouter plan
• Reduce request frequency

Issue 4: Network Connection Failed

Error: connection refused

Solution:

• Check network connection
• Verify OpenRouter service status
• Check proxy settings

Extending

Use Different Models for Different Agents

agents:
  sub:
    - id: "agent-destination"
      type: "destination"
      model: "meta-llama/llama-3.1-70b-instruct"  # Use more powerful model
    
    - id: "agent-food"
      type: "food"
      model: "meta-llama/llama-3.1-8b-instruct"  # Use fast model

Add Request Retry

llm:
  retry:
    max_attempts: 3
    initial_delay: 1s
    backoff_multiplier: 2.0

References

• Main README
• OpenRouter Documentation
• OpenRouter Models
• LLM Documentation
• Quick Start

License

MIT License

Created: 2026-03-23
Example Type: OpenRouter API Integration Demonstration
Code Location: examples/openrouter/main.go:1-409