opentelemetry-semantic-conventions-ai
OpenTelemetry Semantic Conventions Extension for Large Language Models
Decision gist · record as of 2026-08-14
Yes, if you are building observability into a generative AI application and already use OpenTelemetry. It provides the semantic vocabulary needed to instrument LLM calls consistently. If you are not using OpenTelemetry or do not need structured LLM telemetry, it is not applicable.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.9 or later (supports_current); must have opentelemetry-sdk and opentelemetry-semantic-conventions installed as runtime dependencies.
- Low install friction with only two runtime dependencies (opentelemetry-sdk and opentelemetry-semantic-conventions).
- Package is actively maintained with a recent release as of 2026-03-26.
License · maintenance · safety
Apache-2.0 (permissive) — Licensed under Apache-2.0 (permissive), allowing use in commercial and proprietary projects with minimal restrictions.
last release 2026-03-26 (141 days)
0 known vulnerabilities (OSV.dev, 2026-08-14) · 22,238,058 downloads/mo, #981 on PyPI
Alternatives
Verify before relying
pip install opentelemetry-semantic-conventions-ai
from opentelemetry_semantic_conventions_ai import *
# Use AI-specific semantic convention attributes in your spans- Specific list of AI-specific attributes and span conventions provided by this extension
- Integration examples or documentation for common LLM frameworks or providers
- Whether this package includes helper utilities or only defines constants/attributes
What it is and what it does
This package extends the OpenTelemetry semantic conventions standard with attributes and definitions tailored for generative AI applications. It defines structured span attributes for capturing prompt inputs, completion outputs, token counts, model identifiers, and other LLM-specific observability signals that are not covered by the base semantic conventions.
It sits between your instrumentation code and the OpenTelemetry SDK, providing a standardized vocabulary for AI workloads. Rather than inventing your own attribute names, you use the conventions defined here to ensure consistency across your observability stack and compatibility with tools that understand these AI-specific signals.
Use it for
- Instrument LLM API calls to capture prompts, completions, and token usage in structured spans for debugging and cost analysis.
- Monitor generative AI application performance by tracing token consumption and latency across different models and providers.
- Build observability dashboards that correlate LLM behavior (prompt length, completion tokens) with application-level metrics.
- Standardize telemetry collection across multiple gen-AI services so logs and traces use consistent attribute names.
- Audit and track LLM interactions for compliance by recording structured prompt and completion data in spans.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are building observability into a generative AI application and already use OpenTelemetry.
It provides the semantic vocabulary needed to instrument LLM calls consistently. If you are not using OpenTelemetry or do not need structured LLM telemetry, it is not applicable.
Install
opentelemetry-semantic-conventions-ai on PyPI
Before you install
Low install friction with only two runtime dependencies (opentelemetry-sdk and opentelemetry-semantic-conventions). Package is actively maintained with a recent release as of 2026-03-26.
Requires Python 3.9 or later (supports_current); must have opentelemetry-sdk and opentelemetry-semantic-conventions installed as runtime dependencies.
License in practice
Licensed under Apache-2.0 (permissive), allowing use in commercial and proprietary projects with minimal restrictions.
Quickstart
pip install opentelemetry-semantic-conventions-ai
from opentelemetry_semantic_conventions_ai import *
# Use AI-specific semantic convention attributes in your spansVerify before relying
- Specific list of AI-specific attributes and span conventions provided by this extension
- Integration examples or documentation for common LLM frameworks or providers
- Whether this package includes helper utilities or only defines constants/attributes
Package facts
| License | Apache-2.0 permissive |
| Python support | Supports the current Python release <4,>=3.9 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 2 packagesopentelemetry-sdkopentelemetry-semantic-conventions |
| Maintenance | Actively maintained 141 days since the last release |
| First released | |
| Downloads | 22,238,058 / month, #981 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
Evidence: opentelemetry_semantic_conventions_ai-0.5.1-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “gen-ai telemetry instrumentation”
- opentelemetry-semantic-conventions-aiProvides OpenTelemetry semantic conventions and attributes for…
- opentelemetry-instrumentation-voyageaiAutomatically captures and traces Voyage AI API calls (embeddings and…
- vercel-internal-telemetryProvides shared telemetry helpers for internal use across Vercel's…
Give your agent the search over MCP, or paste the wish link into any chat.
More Artificial Intelligence packages
LiteLLM provides a unified Python interface to call 100+ LLM providers (OpenAI, Anthropic, Gemini, Bedrock, Azure, and others) using OpenAI-compatible API format, available as both a Python SDK and a self-hosted AI Gateway proxy server.
Install it if you need to work with multiple LLM providers or want to centralize LLM routing in your organization.
Client library and CLI tool for downloading, uploading, and managing models, datasets, and repositories on the Hugging Face Hub platform.
Install it if you work with Hugging Face Hub models or datasets.
LangChain provides a framework for building agents and LLM-powered applications by composing language models, tools, and memory through a unified API that abstracts over multiple model providers.
hf-xet provides chunk-based deduplication and efficient file transfer for the Hugging Face Hub, enabling faster uploads and downloads of large files with local disk caching.
Tokenizers converts raw text into token sequences for NLP models, with support for training custom vocabularies and using pre-built tokenizers (BPE, WordPiece) optimized for speed via Rust.
Transformers provides a unified framework for loading, fine-tuning, and running state-of-the-art pretrained models across text, vision, audio, video, and multimodal tasks using PyTorch, JAX, or TensorFlow.
Install it if you need to run or train any transformer-based model for NLP, vision, audio, or multimodal tasks.
See also opentelemetry-semantic-conventions · opentelemetry-instrumentation-google-generativeai · opentelemetry-instrumentation-google-genai · aliyun-semantic-conventions · opentelemetry-util-genai · openlit · opentelemetry-instrumentation-openai-v2 · microsoft-agents-a365-observability-core · opentelemetry-instrumentation-litellm · opentelemetry-instrumentation-together