Helicone
Helicone is an AI Gateway and LLM Observability platform that provides a unified API to access 100+ AI models, with intelligent routing and automatic fallbacks. It solves the problem of managing multiple AI providers and tracking costs, latency, and quality across LLM applications. It is designed for AI engineers building and debugging AI-powered applications.
✨ Key features
- Unified API for 100+ AI models with one API key
- Intelligent routing and automatic fallbacks
- One-line integration for logging requests from major providers
- Trace and session inspection for debugging agents and chatbots
- Cost, latency, and quality metrics tracking
- Prompt management and versioning using production data
🎯 Use cases
- Build applications that switch between multiple LLM providers without code changes
- Monitor and debug AI agent conversations and sessions
- Track and analyze cost and latency of LLM API calls
- Version and deploy prompts through the AI Gateway
- Export observability data to PostHog for custom dashboards
📦 Installation
🧰 Requirements: Requires an API key from Helicone (sign up at helicone.ai/signup) and credits. Works with any runtime that supports the OpenAI API.
git clone https://github.com/Helicone/helicone.git
cd docker
cp .env.example .env./helicone-compose.sh helicone up🚀 Usage
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://ai-gateway.helicone.ai",
apiKey: process.env.HELICONE_API_KEY,
});
const response = await client.chat.completions.create({
model: "gpt-4o-mini", // claude-sonnet-4, gemini-2.0-flash or any model from https://www.helicone.ai/models
messages: [{ role: "user", content: "Hello!" }]
});
❓ FAQ
How do I get started with Helicone?
Sign up at helicone.ai/signup to get an API key, then update your OpenAI client's baseURL to https://ai-gateway.helicone.ai and use your Helicone API key.
Can I self-host Helicone?
Yes, Helicone can be self-hosted using the provided docker-compose file. For enterprise workloads, a Helm chart is available by contacting enterprise@helicone.ai.
What integrations are supported?
Helicone supports many inference providers like OpenAI, Anthropic, Gemini, and frameworks like LangChain, Vercel AI SDK, and more. The list may be out of date, so check the docs for the latest.
Is there a free tier?
Yes, Helicone offers a generous monthly free tier of 10k requests per month, no credit card required.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.