Phoenix
AI observability and evaluation from Arize
About
Phoenix from Arize is an open-source observability and evaluation platform for LLM and agent applications. It captures OpenTelemetry-style traces, runs evaluations against the captured data, and visualizes prompt-response pairs, embeddings, and metrics in a local UI. Auto-instrumentation covers OpenAI, Anthropic, Google GenAI, Bedrock, LangGraph, Vercel AI SDK, CrewAI, LlamaIndex, DSPy, and others. Self-host via Docker or use Arize's hosted edition.
Reviews (0)
Leave a Review
No reviews yet. Be the first to review!
Details
- Category
- AI Observability & Evaluation
- Price
- Free
- Platform
- Hybrid
- Difficulty
- Easy (2/5)
- License
- Elastic-2.0
- Added
- Jan 29, 2026
Related Tools
UK AI Security Institute framework for large language model evaluations and benchmarks.
ML experiment tracking, visualization, and collaboration
Open source LLM engineering platform for tracing and analytics
Open-source library for evaluating and tracking LLM applications.
Open-source AI metadata tracker for logging and comparing ML experiments.
Python framework for unit testing and evaluating LLM applications with metrics like G-Eval.