traceloop/openllmetry

What is it?

Key features

• 01Instruments calls to major LLM providers such as OpenAI, Anthropic and Bedrock
• 02Traces queries to vector databases including Chroma, Pinecone, Qdrant and Weaviate
• 03Outputs standard OpenTelemetry data compatible with many observability backends
• 04Offers an SDK that starts tracing with a single init call
• 05Includes integrations for agent frameworks like LangChain, LlamaIndex and CrewAI

Use cases

• Monitoring latency and errors of OpenAI or Anthropic calls in a production LLM service
• Debugging retrieval steps in a RAG pipeline that uses a vector database
• Sending LLM traces to an existing Datadog or Grafana stack
• Adding tracing to an agent built with LangChain or CrewAI

Quick start

1. 1

   Install the Traceloop SDK with pip

   pip install traceloop-sdk
2. 2

   Add Traceloop.init() to your code to start tracing

3. 3

   For local use, disable batch sending to see traces immediately by calling Traceloop.init(disable_batch=True)

Things to know

• !Traces are sent to a backend you configure, so you need an account or self-hosted collector to view them.
• !Versions below v0.49.2 collected telemetry, so upgrade to v0.49.2 or later.

Topics

FAQ

What is openllmetry?

Open-source OpenTelemetry-based observability extensions that trace LLM calls, vector DB queries and agent frameworks in Python apps. OpenLLMetry is a set of extensions built on OpenTelemetry that adds observability to LLM applications. It provides standard OpenTelemetry instrumentations for LLM providers such as OpenAI, Anthropic, Bedrock and Gemini, and for vector databases such as Chroma, Pinecone, Qdrant, Milvus and Weaviate, alongside a Traceloop SDK for quick setup.

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

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

Is openllmetry still maintained?

Actively maintained. The last commit to openllmetry was on Oct 19, 2026.

What can I use openllmetry for?

Monitoring latency and errors of OpenAI or Anthropic calls in a production LLM service Debugging retrieval steps in a RAG pipeline that uses a vector database Sending LLM traces to an existing Datadog or Grafana stack Adding tracing to an agent built with LangChain or CrewAI

How do I install openllmetry?

Install the Traceloop SDK with pip: pip install traceloop-sdk Add Traceloop.init() to your code to start tracing For local use, disable batch sending to see traces immediately by calling Traceloop.init(disable_batch=True)

Where can I get openllmetry?

The source code and installation instructions are on GitHub: https://github.com/traceloop/openllmetry