Helicone/helicone

What is it?

Key features

• 01Logs requests to OpenAI, Anthropic, Gemini and other providers with a one-line integration.
• 02Tracks cost, latency and other metrics, with export to PostHog.
• 03Inspects traces and sessions for agents and chatbots.
• 04Provides a prompt playground and versioned prompt management.
• 05Offers an AI Gateway for 100+ models through one API key, with fallbacks.
• 06Can be self-hosted with Docker Compose.

Use cases

• Debugging and monitoring production LLM applications and agents.
• Tracking token spend and latency across multiple model providers.
• Iterating on prompts and versioning them against production data.
• Running a self-hosted LLM observability stack for data-sensitive teams.

Quick start

1. 1

   Sign up for an API key and add credits

2. 2

   Update the baseURL and API key in your OpenAI client code

3. 3

   Clone the repository

   git clone https://github.com/Helicone/helicone.git
4. 4

   Start the self-hosted services with Docker Compose

   ./helicone-compose.sh helicone up

Things to know

• !Self-hosting requires running five services, including ClickHouse, Supabase and Minio.
• !Manual deployment is not recommended; Docker or Helm is the supported path.
• !Full AI Gateway use requires a Helicone account and API key, and the hosted free tier is capped at 10k requests per month.

Topics

FAQ

What is helicone?

Open-source LLM observability platform and AI gateway for logging, tracing, monitoring and routing requests to 100+ models. Helicone is an open-source platform for monitoring applications built on large language models. It sits between your code and model providers, so requests to OpenAI, Anthropic, Gemini and others can be logged with a one-line change to the base URL or an async logging integration.

Is helicone free to use? What license does it have?

helicone is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.

Is helicone still maintained?

Actively maintained. The last commit to helicone was on Oct 20, 2026.

What can I use helicone for?

Debugging and monitoring production LLM applications and agents. Tracking token spend and latency across multiple model providers. Iterating on prompts and versioning them against production data. Running a self-hosted LLM observability stack for data-sensitive teams.

How do I install helicone?

Sign up for an API key and add credits Update the baseURL and API key in your OpenAI client code Clone the repository: git clone https://github.com/Helicone/helicone.git Start the self-hosted services with Docker Compose: ./helicone-compose.sh helicone up

Where can I get helicone?

The source code and installation instructions are on GitHub: https://github.com/Helicone/helicone