Overview
OpenLIT is an open-source AI engineering platform focused on observability and governance for LLM applications. It offers OpenTelemetry-native SDKs, prompt management, evaluation pipelines, and dashboards to help teams monitor cost, performance and model behavior.
Key features
- Observability: OpenTelemetry-native SDKs and dashboards for tracing LLMs, vectors and GPU metrics.
- Prompt management: versioned prompt hub and templates for consistent deployments.
- Evaluations & guardrails: pipelines for automated evaluation and runtime safety checks.
Use cases
- Add observability and cost tracking to LLM applications.
- Centralize prompt and evaluation workflows across teams.
- Deploy governance and runtime guardrails for model-driven features.
Technical notes
- Multi-language SDKs (Python, TypeScript) with Docker/Kubernetes deployment options.
- Comprehensive documentation and examples at docs.openlit.io.
- Apache-2.0 licensed and actively maintained by the OpenLIT community.