LLM Logger

LLM Logger is a TypeScript/JavaScript library that allows you to easily log your AI API calls to LLM Logger. This package provides direct support for OpenAI, Claude, and Gemini, but you can use it with any LLM by structuring the output yourself.

Installation

To install the llm-logger package, run the following command:

npm install llm-logger

or if you're using yarn:

yarn add llm-logger

Setup

First, import the LLMLogger class from the package:

import { LLMLogger } from 'llm-logger';

Then, create an instance of LLMLogger with your API key:

const logger = new LLMLogger({ apiKey: 'your-api-key-here' });

Usage

OpenAI

To log OpenAI API calls, send the completion object directly as it's returned by the openai library to LLM Logger:

import OpenAI from 'openai';

const openai = new OpenAI({
  apiKey: 'your-openai-api-key',
});

const messages = [{ role: 'user', content: 'Hello, how are you?' }]
const completion = await openai.chat.completions.create({
  model: 'gpt-4o-mini',
  messages,
});

await logger.openaiLog(messages, chatCompletion, ['openai-example'], null);

Claude (Anthropic)

To log Claude API calls:

import Anthropic from '@anthropic-ai/sdk';

const anthropic = new Anthropic({
  apiKey: 'your-anthropic-api-key',
});

const system = 'You are a helpful assistant';
const messages = [{ role: 'user', content: 'Hello, how are you?' }];
const completion = await anthropic.messages.create({
  model: 'claude-3-haiku-20240307',
  system,
  messages,
  max_tokens: 4096,
});

await logger.claudeLog(messages, system, completion, ['claude-example'], null);

Gemini (Google)

To log Gemini API calls:

import { GoogleGenerativeAI } from '@google/generative-ai';

const model = 'gemini-1.5-flash-latest';
const genAI = new GoogleGenerativeAI('your-google-api-key');
const geminiModel = genAI.getGenerativeModel({ model });

const messages = [{ role: 'user', content: 'Hello, how are you?' }];
const result = await geminiModel.generateContent(messages[0].content);
const response = result.response;

await logger.geminiLog(messages, model, geminiResponse, ['gemini-example'], null);

Benefits

Using LLM Logger allows you to:

1. Keep track of all your AI API calls in one place.
2. Analyze usage patterns and optimize your API usage.
3. Debug and improve your AI-powered applications more effectively.

For more information and advanced usage, visit LLM Logger.