LLMMapper

​

Background

LLMMapper acts as a bridge between your OpenAI API calls and Anthropic’s Claude API. It translates your requests on-the-fly, allowing you to use Claude’s capabilities without changing your existing code structure. This integration enables you to leverage the power of Claude models while maintaining compatibility with your OpenAI SDK-based codebase.

​

Using LLMMapper with Helicone Integration

LLMMapper now offers seamless integration with Helicone, allowing you to log and monitor your API calls while using Anthropic’s Claude models through the OpenAI SDK. Here’s how to get started:

​

Step 1: Update your base URL

Instead of using the standard LLMMapper URL, use the Helicone-integrated URL:

const openai = new OpenAI({
  apiKey: process.env.ANTHROPIC_API_KEY,
  baseURL: "https://llmmapper.helicone.ai/oai2ant/v1",
});

​

Step 2: Add your Helicone API Key

To enable logging with Helicone, add your Helicone API key to the request headers:

const openai = new OpenAI({
  apiKey: process.env.ANTHROPIC_API_KEY,
  baseURL: "https://llmmapper.helicone.ai/oai2ant/v1",
  defaultHeaders: {
    "Helicone-Auth": "Bearer YOUR_HELICONE_API_KEY",
  },
});

Replace YOUR_HELICONE_API_KEY with your actual Helicone API key.

​

Step 3: Use as normal

Now you can use the OpenAI SDK with Anthropic’s Claude models as usual, and all your requests will be automatically logged in your Helicone dashboard.

const chatCompletion = await openai.chat.completions.create({
  model: "claude-3-opus-20240229",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Tell me a short joke about programming." },
  ],
  max_tokens: 100,
});

​

Benefits of Helicone Integration

By using LLMMapper with Helicone integration, you get:

• Detailed logging of all your API calls
• Real-time monitoring and analytics
• Cost tracking and optimization insights
• Advanced features like caching and retries

Visit your Helicone dashboard to view logs, analytics, and more for your LLMMapper requests.