Supported Libraries

Every valid instrumentation key for neatlogs.init(instrumentations=[...]), with install commands and version requirements.

Pass any combination of these keys to instrumentations in neatlogs.init(). Install the matching extra first (or the library itself — the extra just pins a known-good version).

neatlogs.init(
    instrumentations=["langchain", "chromadb"],
    api_key="...",
)

Many of these libraries also work with neatlogs.wrap(client) instead of an instrumentations key — wrap() has no import-order rule. See the Python SDK and per-library Integrations for runnable examples.

LLM providers

KeyInstallNotes
openaipip install "neatlogs[openai]"openai>=1.0.0. Also covers AzureOpenAI.
azure_openaipip install "neatlogs[azure-openai]"openai>=1.0.0
azure_ai_inferencepip install "neatlogs[azure-ai-inference]"Azure AI Inference SDK. Required when CrewAI dispatches to Azure.
anthropicpip install "neatlogs[anthropic]"anthropic>=0.75.0
google_genaipip install "neatlogs[google-genai]"google-genai>=1.55.0. Construct the client after init().
google_generativeaiLegacy google.generativeai SDK.
vertex_aipip install "neatlogs[vertex-ai]"Vertex AI via the google-genai SDK in Vertex mode.
vertexaipip install "neatlogs[vertexai]"google-cloud-aiplatform>=1.38.0
bedrockpip install "neatlogs[bedrock]"boto3>=1.42.11
litellmpip install "neatlogs[litellm]"litellm>=1.80.11
cohereCohere SDK.
groqpip install "neatlogs[groq]"groq>=0.37.1
mistralaipip install "neatlogs[mistralai]"mistralai>=1.0.0
togetherTogether AI.
ollamaLocal Ollama models.
replicateReplicate.
openrouterOpenRouter (also via neatlogs.wrap).
huggingface_hubHugging Face Hub inference.
alephalphaAleph Alpha.
watsonxIBM watsonx.
sagemakerAWS SageMaker.
portkeypip install "neatlogs[portkey]"portkey-ai>=1.0.0

Agent frameworks

KeyInstallNotes
langchainpip install "neatlogs[langchain]"Covers LangGraph execution and langchain_openai / langchain_anthropic / etc.
langgraphpip install "neatlogs[langgraph]"langgraph>=1.0.4. Add only for graph-level spans beyond what langchain captures.
crewaipip install "neatlogs[crewai]"crewai>=1.9.3. Also add the provider key matching crewai.LLM(model=...).
openai_agentspip install "neatlogs[openai-agents]"openai-agents>=0.6.5
pydantic_aipip install "neatlogs[pydantic-ai]"pydantic-ai>=0.0.9
dspypip install "neatlogs[dspy]"dspy>=2.6.13 (OpenInference; silent no-op below this — use wrap() instead).
agnopip install "neatlogs[agno]"agno>=2.3.13
google_adkpip install "neatlogs[google-adk]"google-adk>=1.14.1
strandsStrands agents (also neatlogs.strands_hooks(agent)).
autogenpip install "neatlogs[autogen-agentchat]"autogen-agentchat>=0.4.0
haystackpip install "neatlogs[haystack]"haystack-ai>=2.0.0
llamaindexpip install "neatlogs[llama-index]"llama-index>=0.14.10
smolagentspip install "neatlogs[smolagents]"smolagents>=1.0.0
hermespip install "neatlogs[hermes]"hermes-agent>=0.15.1
beeaiBeeAI framework.
pipecatPipecat voice agents.
promptflowMicrosoft Promptflow.

CrewAI routing: "crewai" alone is not enough — CrewAI dispatches via LiteLLM. Add the provider key matching your crewai.LLM(model=...) prefix: gpt-*openai, azure/*azure_ai_inference, gemini/*google_genai, claude-*anthropic.

Vector databases

KeyInstallNotes
chromadbpip install "neatlogs[langchain]" (bundled) or chromadbAuto-instrumented when installed.
pineconeAuto-instrumented when installed.
qdrantAuto-instrumented when installed.
weaviateAuto-instrumented when installed.
milvuspip install "neatlogs[milvus]"pymilvus>=2.4.0,<2.5.0
elasticsearchAuto-instrumented when installed.
redisAuto-instrumented when installed (vector search).
marqoAuto-instrumented when installed.

If you use LangChain retrievers wrapping these (common for RAG), instrumentations=["langchain"] already captures the retrieval spans — a dedicated vector-DB key is only needed when you call the DB client directly.

Other

KeyInstallNotes
mcppip install "neatlogs[mcp]"mcp>=1.0.0. Model Context Protocol tools.
instructorpip install "neatlogs[instructor]"instructor>=1.0.0
guardrailspip install "neatlogs[guardrails]"guardrails-ai>=0.4.0
httpx / requests / aiohttpHTTP-client instrumentation for raw LLM calls.

A library version below its instrumentor's minimum may silently emit no spans (notably dspy < 2.6.13). If a provider isn't traced, check the version, confirm the key is in instrumentations=[], and ensure init() runs before the import. See Python SDK → Auto-instrumentation.

On this page