Introduction to Helicone

What is Helicone?

Helicone is an open-source AI Gateway and LLM observability platform that helps developers build production AI applications with complete visibility and reliability. We solve the hardest problems in production LLM applications: provider outages that break your app, unpredictable costs, and debugging issues that are impossible to reproduce.

AI Gateway

Access 100+ LLM models through a unified OpenAI-compatible API with automatic fallbacks, intelligent routing, and 0% markup

Full Observability

Complete visibility into every request with automatic cost tracking, latency monitoring, and session debugging

Open Source

Apache 2.0 licensed and self-hostable. Your data stays under your control, always accessible

Production Ready

SOC 2 and GDPR compliant with edge deployment for <50ms overhead. Built for your worst day

Key Features

🌐 AI Gateway

Access 100+ AI models with a single API key through our OpenAI-compatible gateway. Switch between OpenAI, Anthropic, Google, Groq, and more by just changing the model name. We handle intelligent routing, automatic fallbacks, and unified billing with 0% markup.

🔍 LLM Observability

Inspect and debug every request with automatic tracking of costs, latency, token usage, and errors. Trace multi-step AI workflows and agent sessions with full visibility into the reasoning chain.

🎮 Prompt Management

Version control your prompts and deploy changes without code updates. Test iterations in production, A/B test versions, and maintain full history of changes.

💰 Cost & Usage Tracking

Understand your unit economics with real-time cost tracking across all providers. Break down spending by user, feature, or any custom dimension. Set budgets and prevent surprise bills.

🎛️ Smart Controls

Caching: Save costs and improve latency with automatic response caching
Rate Limiting: Protect your budget with per-user or global rate limits
Automatic Fallbacks: Stay online even when providers go down
Security: Built-in content moderation and LLM security

Who Uses Helicone?

Startups

Move fast with instant observability. No need to build monitoring infrastructure from scratch.

Enterprises

Self-host for complete data control. SOC 2 compliant with enterprise Helm charts available.

AI Engineers

Debug complex agent workflows. Trace multi-step reasoning with session trees and full context.

Why Helicone?

Best Price Always
We fight for every penny. 0% markup on credits means you pay exactly what providers charge. No hidden fees, no games. Invisible Performance
Your app shouldn’t slow down for observability. Edge deployment keeps us under 50ms overhead. Always. Always Online
Your app stays up, period. Providers fail, we fallback. Rate limits hit, we load balance. We don’t go down. Never Be Surprised
No shock bills. No mystery spikes. See every cost as it happens. We believe in radical transparency. Built for Your Worst Day
When production breaks at 3am and everyone’s panicking, we’re rock solid. Built for when you need us most.

Getting Started

Quick Start

Get your first LLM request logged in under 2 minutes

Quick Start Guide

Send your first request through the AI Gateway and see it logged in real-time

Understand the Platform

Learn how Helicone solves production AI challenges

Platform Overview

Explore the architecture and see how the gateway, observability, and features work together

Explore Features

Go deeper with advanced capabilities

Sessions & Agent Tracing

Debug multi-step AI workflows with full visibility

Prompt Management

Version and deploy prompts without code changes

Gateway Fallbacks

Automatic failover when providers go down

Cost Tracking

Understand your LLM economics by user or feature

Integration Methods

Helicone works in two ways:

AI Gateway (Recommended)
Bring Your Own Keys

Use Credits for Instant Access

Add credits to your Helicone account and access 100+ models without signing up for each provider. We manage the API keys and handle billing at cost (0% markup).Benefits:

Single integration for all providers
Switch models by just changing the name
Automatic observability built-in
No provider rate limits
Unified billing

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://ai-gateway.helicone.ai",
  apiKey: process.env.HELICONE_API_KEY,
});

// Use any model from any provider
const response = await client.chat.completions.create({
  model: "gpt-4o-mini", // or claude-sonnet-4, gemini-2.0-flash, etc.
  messages: [{ role: "user", content: "Hello!" }],
});

Use Your Provider Keys

Configure your own API keys for OpenAI, Anthropic, and other providers at Provider Keys. You manage billing directly with providers while getting full Helicone observability.When to use:

You have existing provider relationships
You need provider-specific features
You want direct provider billing

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://oai.helicone.ai/v1",
  apiKey: process.env.OPENAI_API_KEY,
  defaultHeaders: {
    "Helicone-Auth": `Bearer ${process.env.HELICONE_API_KEY}`,
  },
});

Supported Providers

Helicone integrates with 100+ LLM providers including:

OpenAI

GPT-4o, GPT-4o-mini, o1, o3

Anthropic

Claude Sonnet 4, Opus, Haiku

Google

Gemini 2.0, Gemini Pro, Vertex AI

Groq

Llama, Mixtral, Gemma

AWS Bedrock

Claude, Llama, Titan

Azure OpenAI

All Azure-hosted models

Together AI

Open-source models

And More

Mistral, DeepSeek, Perplexity, X.AI, Fireworks, and more

See the complete list of supported models.

Framework Integrations

Helicone works seamlessly with popular AI frameworks:

LangChain (JS/TS & Python) - Use AI Gateway for unified provider access
LlamaIndex (Python) - Full observability for data applications
LangGraph (Python) - Track stateful multi-actor workflows
Vercel AI SDK (JS/TS) - Build AI-powered applications
CrewAI (Python) - Monitor role-playing AI agents

See all integrations in our docs.

Open Source & Self-Hosting

Helicone is Apache 2.0 licensed and fully open source. Self-host for complete data ownership and control.

Docker Setup

Get started with Docker Compose in minutes

Helm Chart

Production-ready Kubernetes deployment (contact [email protected])

GitHub: github.com/helicone/helicone
⭐ Star us if you find Helicone useful!

Community & Support

Discord Community

Join 2000+ developers building with Helicone

Documentation

Comprehensive guides and API references

Email Support

[email protected] for questions and feedback

GitHub Issues

Report bugs and request features

What’s Next?

Ready to get started? Follow our quick start guide to log your first request in under 2 minutes.

Quick Start Guide

Send your first LLM request through Helicone and see it logged in real-time

Get Started

AI Gateway

Observability

Prompt Management

Evaluation & Testing

Features

Self-Hosting

Integrations

​What is Helicone?

AI Gateway

Full Observability

Open Source

Production Ready

​Key Features

​🌐 AI Gateway

​🔍 LLM Observability

​🎮 Prompt Management

​💰 Cost & Usage Tracking

​🎛️ Smart Controls

​Who Uses Helicone?

Startups

Enterprises

AI Engineers

​Why Helicone?

​Getting Started

Quick Start Guide

Platform Overview

Sessions & Agent Tracing

Prompt Management

Gateway Fallbacks

Cost Tracking

​Integration Methods

​Use Credits for Instant Access

​Use Your Provider Keys

​Supported Providers

OpenAI

Anthropic

Google

Groq

AWS Bedrock

Azure OpenAI

Together AI

And More

​Framework Integrations

​Open Source & Self-Hosting

Docker Setup

Helm Chart

​Community & Support

Discord Community

Documentation

Email Support

GitHub Issues

​What’s Next?

Quick Start Guide

Build docs developers (and LLMs) love

What is Helicone?

Key Features

🌐 AI Gateway

🔍 LLM Observability

🎮 Prompt Management

💰 Cost & Usage Tracking

🎛️ Smart Controls

Who Uses Helicone?

Why Helicone?

Getting Started

Integration Methods

Use Credits for Instant Access

Use Your Provider Keys

Supported Providers

Framework Integrations

Open Source & Self-Hosting

Community & Support

What’s Next?