# Semantic Conventions GenAI Reference Implementations Validates [OpenTelemetry Semantic Conventions for Generative AI](https://opentelemetry.io/docs/specs/semconv/gen-ai/) against real LLM client libraries in Python, showing which libraries support which attributes. Each library under [scenarios/](scenarios/) contains a small reference implementation (`scenario.py`) that exercises the SDK against a deterministic local mock server and emits OpenTelemetry spans, metrics, and logs, plus a `conformance.yaml` saying how to run it. The [conformance runner](https://github.com/open-telemetry/semantic-conventions-conformance) validates the captured telemetry against the semantic conventions in [../model/](../model/) using [OTel Weaver](https://github.com/open-telemetry/weaver) and writes the per-library results to `scenarios//data.json`, which feed the status reports below. See [CONTRIBUTING.md](CONTRIBUTING.md) for how to run scenarios and add new libraries. ## Reports Generated from committed `scenarios/*/data.json` files. Do not edit this section by hand. Run `uv run update-reports` to regenerate. ### Spans | Span | Libraries | | --- | --- | | [A2A Client](reports/a2a-client-span.md) | a2a-python | | [A2A Server](reports/a2a-server-span.md) | a2a-python | | [MCP Client](reports/mcp-client-span.md) | mcp | | [MCP Server](reports/mcp-server-span.md) | mcp | | [Create Agent](reports/create-agent-span.md) | anthropic, aws-bedrock-agent, azure-ai-foundry, google-genai, mistralai, openai-assistants | | [Invoke Agent Client](reports/invoke-agent-client-span.md) | aws-bedrock-agent, azure-ai-foundry, google-genai, openai-assistants | | [Invoke Agent Internal](reports/invoke-agent-internal-span.md) | adk_a2a, agent-framework, autogen, crewai, google-adk, langchain, openai-agents, pydantic-ai | | [Plan](reports/plan-span.md) | crewai, langchain | | [Inference](reports/inference-span.md) | adk_a2a, agent-framework, anthropic, aws-bedrock, azure-ai-inference, azure-openai, claude-agent-sdk, cohere, google-genai, groq, litellm, mistralai, openai, vertexai | | [Realtime Inference](reports/realtime-inference-span.md) | google-genai, openai | | [User Speech](reports/user-speech-span.md) | google-genai, openai | | [Embeddings](reports/embeddings-span.md) | aws-bedrock, azure-ai-inference, azure-openai, cohere, google-genai, litellm, mistralai, openai | | [Retrieval](reports/retrieval-span.md) | haystack, langchain, llamaindex | | [Fetch Response](reports/fetch-response-span.md) | openai | | [Memory](reports/memory-span.md) | aws-bedrock-agentcore, google-adk | | [Execute Tool](reports/execute-tool-span.md) | agent-framework, autogen, crewai, google-adk, google-genai, langchain, llamaindex, openai-agents, openai-assistants, pydantic-ai, vertexai | ### Events | Event | Libraries | | --- | --- | | [Realtime Session Started](reports/gen-ai-client-realtime-session-started-event.md) | google-genai, openai | | [Realtime Session Ended](reports/gen-ai-client-realtime-session-ended-event.md) | google-genai, openai | | [Client Embeddings Exception](reports/gen-ai-client-embeddings-exception-event.md) | openai | | [Client Fetch Response Exception](reports/gen-ai-client-fetch-response-exception-event.md) | openai | | [Client Inference Exception](reports/gen-ai-client-inference-exception-event.md) | openai | | [Inference Operation Details](reports/gen-ai-client-inference-operation-details-event.md) | anthropic, aws-bedrock, azure-ai-inference, cohere, google-genai, groq, litellm, mistralai, openai, vertexai | | [Client Invoke Agent Exception](reports/gen-ai-client-invoke-agent-exception-event.md) | aws-bedrock-agent, azure-ai-foundry | | [Evaluation Result](reports/gen-ai-evaluation-result-event.md) | azure-ai-evaluation, deepeval, dspy | | [Execute Tool Exception](reports/gen-ai-execute-tool-exception-event.md) | openai-agents | | [Invoke Agent Exception](reports/gen-ai-invoke-agent-exception-event.md) | openai-agents | ### Entities | Entity | Libraries | | --- | --- | | [Main Agent](reports/gen-ai-main-agent-entity.md) | adk_a2a | ### Metrics | Metric | Libraries | | --- | --- | | [Client Inference Usage Input Tokens](reports/gen-ai-client-inference-usage-input-tokens-metric.md) | anthropic, google-genai, openai | | [Client Inference Usage Output Tokens](reports/gen-ai-client-inference-usage-output-tokens-metric.md) | anthropic, google-genai, openai | | [Client Inference Cache Read Input Tokens](reports/gen-ai-client-inference-usage-cache-read-input-tokens-metric.md) | anthropic, google-genai, openai | | [Client Inference Cache Write Input Tokens](reports/gen-ai-client-inference-usage-cache-write-input-tokens-metric.md) | anthropic, openai | | [Client Inference Reasoning Output Tokens](reports/gen-ai-client-inference-usage-reasoning-output-tokens-metric.md) | google-genai, openai | | [Client Inference Operation Input Tokens](reports/gen-ai-client-inference-operation-input-tokens-metric.md) | anthropic, azure-openai, google-genai, openai | | [Client Inference Operation Output Tokens](reports/gen-ai-client-inference-operation-output-tokens-metric.md) | anthropic, azure-openai, google-genai, openai | | [Client Operation Duration](reports/gen-ai-client-operation-duration-metric.md) | adk_a2a | | [Client Inference Duration](reports/gen-ai-client-inference-duration-metric.md) | agent-framework, anthropic, groq | | [Client Embeddings Duration](reports/gen-ai-client-embeddings-duration-metric.md) | aws-bedrock, azure-ai-inference, azure-openai, cohere, google-genai, litellm, mistralai, openai | | [Client Embeddings Operation Input Tokens](reports/gen-ai-client-embeddings-operation-input-tokens-metric.md) | aws-bedrock, azure-ai-inference, azure-openai, cohere, litellm, mistralai, openai | | [Execute Tool Duration](reports/gen-ai-execute-tool-duration-metric.md) | google-adk | | [Retrieval Duration](reports/gen-ai-retrieval-duration-metric.md) | haystack, langchain, llamaindex | | [Client Invoke Agent Duration](reports/gen-ai-client-invoke-agent-duration-metric.md) | aws-bedrock-agent, azure-ai-foundry, google-genai, openai-assistants | | [Invoke Agent Inference Calls](reports/gen-ai-invoke-agent-inference-calls-metric.md) | adk_a2a, google-adk | | [Invoke Agent Tool Calls](reports/gen-ai-invoke-agent-tool-calls-metric.md) | adk_a2a, google-adk | | [MCP Client Operation Duration](reports/mcp-client-operation-duration-metric.md) | mcp | | [MCP Server Operation Duration](reports/mcp-server-operation-duration-metric.md) | mcp |