Arize-ai/openinference

OpenTelemetry Instrumentation for AI Observability

What it solves

OpenInference provides a standardized way to trace and monitor AI applications. It solves the problem of fragmented observability in LLM-based apps by creating a common set of conventions for recording how LLMs, vector stores, and external tools are invoked within a complex application workflow.

How it works

It functions as a complementary set of conventions and plugins for OpenTelemetry. It defines a specification for recording AI-specific metadata (semantic conventions) and provides a wide array of instrumentation libraries for popular ML SDKs and frameworks (such as OpenAI, LangChain, LlamaIndex, and Anthropic). These instrumentations allow developers to capture detailed execution traces that can be sent to any OpenTelemetry-compatible backend, including Arize Phoenix and Arize AX.

Who it’s for

AI engineers and developers building complex LLM applications, agents, and RAG systems who need deep visibility into their application's internal logic, tool usage, and model calls.

Highlights

  • OpenTelemetry Integration: Built to work seamlessly with the OpenTelemetry ecosystem.
  • Extensive Framework Support: Offers dedicated instrumentation for dozens of popular AI libraries, including DSPy, CrewAI, and PydanticAI.
  • Transport Agnostic: The specification is designed to be independent of specific file formats or transport methods.
  • Unified Data: Includes span processors to normalize data from other instrumentation libraries like OpenLIT and OpenLLMetry.

Related

  • Project
    traceloop/openllmetry

    An open-source observability framework built on OpenTelemetry that provides complete tracing and monitoring for LLM applications, providers, and vector databases.

  • Project
    Arize-ai/phoenix

    An open-source AI observability platform for tracing, evaluating, and troubleshooting LLM applications through OpenTelemetry-based instrumentation.

  • Project
    openlit/openlit

    An open-source observability and evaluation platform for AI agents that uses OpenTelemetry to trace, debug, and optimize LLM applications.

  • Project
    agentjido/req_llm

    ReqLLM is an Elixir library that unifies calls to dozens of LLM providers (OpenAI, Anthropic, Azure, Google Gemini, etc.) behind a single, high‑level API. It offers functions like `generate_text/3`, `stream_text/3`, and `generate_object/4` that work across 1,200+ models, handles streaming via Finch, validates structured outputs, tracks token usage and estimated cost, and provides built‑in Telemetry. Keys are managed automatically from env files or in‑memory storage, and the library includes extensive provider guides and a Discord community for support.

  • Dispatch
    Introducing opensmith: A Local-First, Open-Source Alternative to LangSmith

    opensmith provides a local-first LLM pipeline tracer that eliminates the need for cloud accounts and hosted services, offering a zero-setup observability tool for Python developers.