arize-phoenix
AI Observability and Evaluation
Decision gist · record as of 2026-08-14
Yes, if you are building or maintaining LLM applications and need observability. The platform is actively maintained, has strong community adoption (11052 stars), and offers a comprehensive feature set for tracing, evaluation, and experimentation. The 51 dependencies and unclear license (Elastic-2.0) warrant review before production use; verify license terms and whether your deployment model (local, cloud, or self-hosted) aligns with your requirements.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires Python 3.10 or later.
- The platform runs as a server process; tracing requires OpenTelemetry instrumentation setup in your application code.
- Low install friction with a pure-Python wheel.
License · maintenance · safety
Elastic-2.0 (unclear) — Licensed under Elastic-2.0, which is not a standard SPDX identifier and its treatment is unclear. Verify the license terms before use in proprietary or commercial contexts, as Elastic licenses can carry specific restrictions.
last release 2026-08-13 (1 days) · last repo commit 2026-08-14 · 11,052 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 2,331,841 downloads/mo, #3,132 on PyPI
Alternatives
Verify before relying
pip install arize-phoenix
phoenix serve
# In your application:
from openinference.instrumentation.openai import OpenAIInstrumentation
from arize_phoenix.otel import register
register()
OpenAIInstrumentation().instrument()- Whether Elastic-2.0 is a typo or variant of a known license; clarify commercial use restrictions
- Whether the 51 runtime dependencies can be selectively installed or if all are required
- Whether the platform requires external services or can run fully self-contained locally
What it is and what it does
Phoenix is a web-based observability and evaluation platform for LLM applications. It captures execution traces using OpenTelemetry instrumentation, stores them in a local or remote database, and provides a UI for inspecting traces, running LLM-based evaluations, managing versioned datasets, and experimenting with prompts and model parameters. The platform is vendor-agnostic and integrates with popular frameworks like LangGraph, LlamaIndex, OpenAI Agents, and Claude Agent SDK, as well as major LLM providers.
You run Phoenix as a server (locally via `phoenix serve`, in Docker, or in Kubernetes) and instrument your application to send traces to it. The platform then lets you replay traces, benchmark performance, version and organize test datasets, and iterate on prompts and retrieval strategies. It also includes an MCP server endpoint for integration with coding agents like Claude Code and Cursor.
Use it for
- Trace and inspect LLM application execution to debug unexpected outputs or performance issues
- Run LLM-based evaluations on application responses and retrieval quality across versioned datasets
- Organize and version datasets for experimentation, evaluation, and fine-tuning workflows
- Compare prompt variations and model parameters side-by-side using the playground and experiment tracking
- Deploy a self-hosted observability backend for teams that need on-premises or air-gapped LLM monitoring
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you are building or maintaining LLM applications and need observability.
The platform is actively maintained, has strong community adoption (11052 stars), and offers a comprehensive feature set for tracing, evaluation, and experimentation. The 51 dependencies and unclear license (Elastic-2.0) warrant review before production use; verify license terms and whether your deployment model (local, cloud, or self-hosted) aligns with your requirements.
Install
arize-phoenix on PyPI
Before you install
Low install friction with a pure-Python wheel. Actively maintained with a release 1 day old and 11052 GitHub stars. Depends on 51 runtime packages including FastAPI, OpenTelemetry, and specialized Phoenix sub-packages, which may add setup complexity despite the wheel distribution.
Requires Python 3.10 or later. The platform runs as a server process; tracing requires OpenTelemetry instrumentation setup in your application code.
License in practice
Licensed under Elastic-2.0, which is not a standard SPDX identifier and its treatment is unclear. Verify the license terms before use in proprietary or commercial contexts, as Elastic licenses can carry specific restrictions.
Quickstart
pip install arize-phoenix
phoenix serve
# In your application:
from openinference.instrumentation.openai import OpenAIInstrumentation
from arize_phoenix.otel import register
register()
OpenAIInstrumentation().instrument()Verify before relying
- Whether Elastic-2.0 is a typo or variant of a known license; clarify commercial use restrictions
- Whether the 51 runtime dependencies can be selectively installed or if all are required
- Whether the platform requires external services or can run fully self-contained locally
Package facts
| License | Elastic-2.0 unclear |
| Python support | Supports the current Python release <3.15,>=3.10 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 51 packagesaioitertoolsaiosqlitealembicarize-phoenix-clientarize-phoenix-evalsarize-phoenix-otelarize-phoenix-sqleanauthlibbashkitcachetoolsemail-validatorfastapifastmcp-slimgrpc-interceptorgrpciohttpxjinja2jmespathjoserfcjsonpath-ngjsonschemaldap3numpyopeninference-instrumentation-openaiopeninference-instrumentationopeninference-semantic-conventionsopentelemetry-exporter-otlpopentelemetry-protoopentelemetry-sdkorjson |
| Maintenance | Actively maintained 1 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 2,331,841 / month, #3,132 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Programming Language :: PythonProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14 |
Evidence: arize_phoenix-20.2.0-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 › “experiment tracking for llms”
- arize-phoenixPhoenix is an open-source AI observability platform that traces LLM…
- tracemlTraceML tracks metrics, parameters, artifacts, and data references…
- sagemaker-experimentsTracks machine learning experiments, trials, and trial components in…
Give your agent the search over MCP, or paste the wish link into any chat.
More Monitoring packages
Wraps any iterable to display a real-time progress bar in the terminal or Jupyter notebook, showing iteration count, elapsed time, and estimated time remaining.
Provides generated Python code for OpenTelemetry semantic conventions, enabling standardized attribute naming and constant definitions for instrumentation and telemetry collection.
Install it if you are using OpenTelemetry and want to follow semantic conventions correctly.
Provides the reference implementation of the OpenTelemetry API for collecting and exporting traces, metrics, and logs from Python applications.
Provides the abstract API and interfaces for OpenTelemetry instrumentation in Python, defining how to emit traces, metrics, and logs without tying code to a specific SDK implementation.
Exports OpenTelemetry observability data to an OpenTelemetry Collector using Protobuf-encoded messages over HTTP.
Install it if you are using OpenTelemetry in Python and need to send data to a Collector over HTTP.
Provides automatic instrumentation commands and programmatic APIs to inject distributed tracing into Python applications without code changes, detecting and instrumenting packages used by your program.
Install it if you need distributed tracing without code changes and have compatible instrumented packages in your environment.
See also arize-phoenix-client · arize-phoenix-evals · arize-phoenix-otel · opik · arize · openinference-instrumentation-litellm · openinference-instrumentation-google-genai · openinference-instrumentation-haystack · openinference-instrumentation-anthropic · openinference-instrumentation-llama-index