LibreAI / Agents / openlit
// agents Β· openlit/openlit

openlit

Open-source observability for AI agents: trace, evaluate, guard, and improve.

πŸ’Ž Hidden gemβ˜… 2.8k starsβ‘‚ 418 forksApache-2.0updated 2026-10
β–¦ visual β€” this project's README ships no embeddable screenshot,
so here's a placeholder. The guide below is complete regardless.
// project facts

Project facts

openlit is an open-source agents project by openlit (openlit/openlit on GitHub). It has 2,814 GitHub stars and 418 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 1 October 2026.

Repository
openlit/openlit
GitHub stars
2,814
Forks
418
License
Apache-2.0
Last updated on GitHub
1 October 2026
Category
Agents
Listing
Hidden gem β€” a shorter page for a lesser-known project
Official site
Official site
// how it works

How it works

  1. Instrument your agent

    Add OpenLIT tracing to your agent code.

  2. Collect traces

    See every model call, tool use, and decision.

  3. Evaluate quality

    Score agent behavior against your standards.

  4. Guard and improve

    Add guardrails and iterate on weak spots.

// features

Why people love it

OTEL-based tracing

Standard observability under the hood.

Agent evaluation

Score quality, not just latency.

Guardrails

Catch bad behavior before users see it.

Harness engineering

Improve everything around the model.

ClickHouse backend

Scales to serious trace volumes.

Multi-agent support

Works with Claude Code, Codex, and more.

Advertisement
// run it

Run it yourself

  1. Step 1 β€” Clone the repo

    Download the source and move into the project folder.

    git clone https://github.com/openlit/openlit.git
    cd openlit
  2. Step 2 β€” Check the README

    Every project documents its own setup; start with install requirements.

    cat README.md | head -60
  3. Step 3 β€” Run it

    Follow the project’s launch command β€” Docker, pip, npm, or a binary release.

    # see README for the exact launch command
// use cases

What you can build

Advertisement
// faq

Questions, answered

Who maintains OpenLIT?

It is an open-source project at openlit/openlit with 418 forks.

Is OpenLIT free for commercial use?

Yes. Apache-2.0 licensed.

How is it different from Langfuse?

It is OTEL-native and frames itself around harness engineering.

Can I self-host it?

Yes. It is designed to run on your own stack.

Which frameworks are supported?

Claude Code, Codex, and others; check the docs for integrations.

// keep exploring

Similar tools

AI Agents

agentops

Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI …

β˜… 5.9k starsMIT
View guide β†’
AI Agents

mcp-use

The fullstack MCP framework to develop MCP Apps for ChatGPT / Claude & MCP Servers for AI Agents.

β˜… 10.7k starsMIT
View guide β†’
AI Agents

Langgraph

Build resilient agents

β˜… 42.7k starsMIT
View guide β†’
AI Agents

opik

Open-source LLM observability: trace, evaluate and monitor apps and agents

β˜… 22.5k starsApache-2.0
View guide β†’
AI Agents

phoenix

Observability and evaluation for AI apps β€” see what your agents do.

β˜… 11.7k starsNOASSERTION
View guide β†’
AI Agents

helicone

Open-source LLM observability: monitor, evaluate, and experiment.

β˜… 6.2k starsApache-2.0
View guide β†’
// alternatives

Alternatives in Agents

The most-starred other Agents projects in the LibreAI catalogue β€” the realistic shortlist if openlit is not the right fit.

Get openlit

Free and open β€” grab the source, star the repo, and start creating.