Logfire
ActiveDescription
AI observability platform for production LLM and agent systems by the Pydantic team. Provides real-time monitoring, tracing, and debugging capabilities.
Key Features
- Python-centric observability with rich object display, event-loop telemetry, and code profiling
- SQL-based querying of all trace data, compatible with existing BI tools and libraries
- OpenTelemetry-native with full support for traces, metrics, and logs across any language
- Built-in Pydantic model integration for validation analytics and data flow visibility
- One-click instrumentation for FastAPI, SQLAlchemy, HTTPX, and other popular Python packages
- Simple and powerful dashboard designed for entire engineering team adoption
Use Cases
💡 Monitoring LLM API calls and agent workflows in production with real-time tracing
💡 Debugging Pydantic model validation failures with built-in analytics dashboards
💡 Profiling Python application performance including database queries and HTTP requests
💡 Centralizing observability across microservices using OpenTelemetry-compatible instrumentation
💡 Analyzing token usage and latency patterns in AI agent pipelines
Tags
Categories
Quick Start
Install with 'pip install logfire', authenticate with 'logfire auth', then add logfire.configure() to your app. For FastAPI integration, call logfire.instrument_fastapi(app). Use logfire.span() for manual tracing or logfire.info() for logging. Query your data with standard SQL in the dashboard.