Python SDK API Reference
Complete version-specific reference documentation. It includes class and method
signatures, parameter descriptions, return types, code examples, etc.
What You Can Do
- LLM Observability: Trace and evaluate LLM applications using OpenTelemetry-based instrumentation
- Experimentation: Create datasets, run experiments, and track model performance
- Evaluation: Analyze and improve model quality with automated evaluation frameworks
- ML Monitoring: Log predictions and actuals for traditional ML models (classification, regression, ranking, etc.)
Choose Your Version
Version 8 (Pre-release)
Comprehensive SDK for the entire Arize platform. Includes LLM tracing, experiments, datasets, and ML monitoring with enhanced customization and modern interfaces. Recommended for new projects.
Version 7
Stable release supporting LLM tracing and traditional ML monitoring.
Version Support
The following table shows the support timeline for each major version of the Arize Python SDK:
We recommend upgrading to the latest version to receive the newest features, improvements, and security updates.