Apache-2.0 ยท Tools

OpenLIT

OpenTelemetry-native open-source AI engineering platform for LLM observability, evaluations, guardrails, prompts, and GPU monitoring.

2.5K stars 0.3K forks Apache-2.0 license 2026-06-10 verified
bash
$# OpenLIT
$pip install openlit
$npx openlit --help
Open source
Overview

What is OpenLIT?

OpenLIT is an open-source AI engineering platform for observability, evaluations, guardrails, prompt management, vault workflows, playgrounds, and GPU monitoring. It integrates with many LLM providers, vector databases, and agent frameworks.

OpenTelemetry-native observability

OpenLIT focuses on AI observability through OpenTelemetry-native tracing and monitoring.

Teams can connect agent behavior to existing observability systems instead of creating isolated AI dashboards.

Evaluation and guardrails

The platform includes evaluations and guardrail workflows.

Operational visibility is stronger when paired with repeatable quality and safety checks.

Broad integration surface

OpenLIT describes integrations across LLM providers, vector databases, agent frameworks, and GPUs.

Agent stacks are heterogeneous, so observability tools need broad coverage.
Use cases

What teams use it for

Agent tracing

Trace model calls, tools, latency, and failures across production agent workflows.

Evaluation monitoring

Connect evaluations and guardrails to ongoing LLM application operations.

AI platform operations

Monitor provider usage, GPU behavior, and prompt workflows in one engineering platform.

Ecosystem

Tags & capabilities

toolopen sourceautomationworkflowopen source
Comparison

How it stacks up

When to choose OpenLIT

Compare it with nearby tools by looking at hosting model, integration surface, license, and whether the official docs show the workflow you need.

FAQ

Questions

Is OpenLIT open source?

Yes. The GitHub repository is listed under the Apache-2.0 license.

How does OpenLIT fit with MLflow or Langfuse?

OpenLIT is especially interesting for teams that want OpenTelemetry-native observability and operational monitoring around LLM and agent systems.