All updatesOctober 22, 2025

Zero-config observability for every framework

You can now instrument OpenAI, Anthropic, LangChain, CrewAI, and a dozen vector databases on neatlogs with a single init call, no decorators, no wrappers.

Zero-config observability for every framework

Instrument your AI stack with a single configuration option

Capture LLM calls, tool invocations, retrievals, and more without adding manual wrappers.

Pass the libraries you use to instrumentations and Neatlogs automatically instruments supported frameworks and SDKs. Just make sure neatlogs.init() runs before importing any instrumented library.

neatlogs.init(
    api_key=os.environ["NEATLOGS_API_KEY"],
    workflow_name="rag-pipeline",
    instrumentations=["openai", "langchain", "chromadb"],
)
See all supported libraries →
  • FeatureAnthropic, azure_ai_inference, bedrock, google_genai, and litellm added as instrumentation keys
  • FeatureCrewAI support via neatlogs.bind_templates() and neatlogs.register_crewai_task()
  • Featurelanggraph instrumentation key for graph-level span capture
  • FeatureMCP_TOOL span kind for tools invoked via Model Context Protocol
  • Featureinstructor and guardrails added as supported instrumentation libraries
  • EnhancementSpan deduplication handles overlapping spans when two instrumentation layers cover the same call
All updates

Browse the full changelog feed

Read the docs

API reference, guides, and examples