Skip to main content
hc
Submit
mcpSource-backedReview first Safety ✓ Privacy ✓

Arize Phoenix MCP Server for Claude

Inspect LLM traces and spans, manage prompts, explore datasets, and review evaluation experiments from Claude — with the official Arize Phoenix MCP server, built into the open-source Phoenix AI observability platform.

by Arize AI·added 2026-06-18·
Claude CodeCodexCursorClaude Desktop
HarnessClaude CodeCodexCursorClaude Desktop
Review first — review before installing

Open the source and read safety notes before installing.

Safety notes

  • The `write` and upsert tools can create or modify prompts and add dataset examples — review Claude's proposed changes before confirming.
  • For self-hosted deployments, ensure your Phoenix instance is accessible from the machine running the MCP server.

Privacy notes

  • LLM traces (including prompts and completions), evaluation annotations, session data, datasets, and experiment results from your Phoenix workspace are surfaced in Claude's context.
  • Your `PHOENIX_API_KEY` is passed as an environment variable — store it in a secrets manager rather than plaintext shell config.

Prerequisites

  • An Arize Phoenix account (cloud at app.phoenix.arize.com) or self-hosted Phoenix instance.
  • A Phoenix API key from your account settings.
  • Node.js 18+ for `npx`.
  • An MCP client such as Claude Code or Claude Desktop.

Schema details

Install type
cli
Troubleshooting
No
Source repository stats
Scope
Source repo
Collection metadata
Estimated setup
5 minutes
Difficulty
beginner
Tool listing metadata
Full copyable content
{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": ["-y", "@arizeai/phoenix-mcp@latest"],
      "env": {
        "PHOENIX_HOST": "https://app.phoenix.arize.com",
        "PHOENIX_API_KEY": "your-api-key"
      }
    }
  }
}

About this resource

Overview

The Arize Phoenix MCP Server is the official Model Context Protocol server for Arize Phoenix, the open-source AI observability platform. It connects Claude directly to your Phoenix workspace — traces, spans, prompts, datasets, sessions, and evaluation experiments — through a comprehensive set of tools for LLM debugging and evaluation.

Phoenix supports both cloud (app.phoenix.arize.com) and self-hosted deployments, so trace data can stay entirely within your own infrastructure.

Key capabilities

  • Traces & spans — retrieve LLM execution traces, individual spans, and annotations.
  • Projects — list projects, get project-level statistics.
  • Sessions — explore conversation sessions and multi-turn flows.
  • Prompt management — list prompts, retrieve versions, and upsert new prompt definitions.
  • Datasets — discover datasets and add new examples.
  • Experiments — access experiment runs and LLM-analyzed evaluation results.
  • Annotation configs — list available evaluation configurations and their schemas.

How it compares

Server Traces & spans Prompt library Datasets Experiments Auth
Arize Phoenix MCP Yes Yes Yes Yes API key
Opik MCP Yes Yes Yes Yes API key
LangSmith MCP Yes Yes Yes Yes API key
Weights & Biases MCP Yes Limited Yes Yes API key

Arize Phoenix is fully open-source (Apache-2.0) with a first-class self-hosted deployment path, making it well-suited for teams with strict data residency requirements.

Installation

Claude Code

claude mcp add phoenix \
  -e PHOENIX_HOST=https://app.phoenix.arize.com \
  -e PHOENIX_API_KEY=your-api-key \
  -- npx -y @arizeai/phoenix-mcp@latest

Get your API key from Phoenix account settings → API Keys.

Claude Desktop

{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": ["-y", "@arizeai/phoenix-mcp@latest"],
      "env": {
        "PHOENIX_HOST": "https://app.phoenix.arize.com",
        "PHOENIX_API_KEY": "your-api-key"
      }
    }
  }
}

Self-hosted Phoenix

Set PHOENIX_HOST=http://localhost:6006 (or your instance URL) instead of the cloud endpoint.

Optional configuration

Set PHOENIX_PROJECT=my-project to default tool calls to a specific project.

Requirements

  • An Arize Phoenix account (cloud or self-hosted).
  • A Phoenix API key.
  • Node.js 18+ (for npx).
  • An MCP client (Claude Code or Claude Desktop).

Security

  • API key scoped to your Phoenix workspace.
  • Self-hosted deployment keeps all trace and evaluation data within your infrastructure.

Source Verification Notes

Verified on 2026-06-18:

  • GitHub repo Arize-ai/phoenix contains the MCP package at js/packages/phoenix-mcp/. The README documents the npx @arizeai/phoenix-mcp@latest install, PHOENIX_HOST / PHOENIX_API_KEY env vars, and the full tool surface (prompts, projects, traces, sessions, datasets, experiments).
  • Phoenix documentation at arize.com/docs/phoenix covers the observability platform, tracing concepts, and evaluation workflows.
#arize-phoenix#llm-observability#tracing#evaluation#evals#prompt-management#ai-observability

Source citations

Add this badge to your README

Show that Arize Phoenix MCP Server for Claude is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.

Listed on HeyClaude
[![Listed on HeyClaude](https://heyclau.de/badge/mcp/arize-phoenix-mcp-server.svg)](https://heyclau.de/entry/mcp/arize-phoenix-mcp-server)

How it compares

Arize Phoenix MCP Server for Claude side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.

FieldArize Phoenix MCP Server for Claude

Inspect LLM traces and spans, manage prompts, explore datasets, and review evaluation experiments from Claude — with the official Arize Phoenix MCP server, built into the open-source Phoenix AI observability platform.

Open dossier 		Opik MCP Server for Claude

Debug, evaluate, and monitor LLM applications from Claude — read traces and spans, score outputs, save prompts, run evaluation experiments, and query project metrics — with the official Opik MCP server by Comet.

Open dossier 		LangSmith MCP Server for Claude

Connect Claude to LangSmith — retrieve conversation threads and traces, fetch and push prompts, browse evaluation datasets and experiments, and access billing usage — with the official LangSmith Model Context Protocol server from LangChain.

Open dossier 		OpenTelemetry MCP Server for Claude

Analyze distributed traces and LLM observability data from Claude — search traces and spans, find errors, list services, analyze LLM token usage, identify slow LLM operations, and discover AI model usage patterns — with the OpenTelemetry MCP server supporting Jaeger, Grafana Tempo, and Traceloop backends.

Open dossier
Trust
Install riskReview firstReview firstReview firstReview first
Notes Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓ Safety ✓ Privacy ✓
Categorymcpmcpmcpmcp
Sourcesource-backedsource-backedsource-backedsource-backed
AuthorArize AICometLangChainTraceloop
Added2026-06-182026-06-182026-06-182026-06-18
Platforms
Claude CodeCodexCursorClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Claude CodeClaude Desktop
Source repo————
Safety notes✓The `write` and upsert tools can create or modify prompts and add dataset examples — review Claude's proposed changes before confirming. For self-hosted deployments, ensure your Phoenix instance is accessible from the machine running the MCP server.✓The `write` tool can create traces, spans, scores, and prompts in your Opik workspace — review any write operations before confirming. The `run_experiment` tool executes evaluation experiments end-to-end, which may incur LLM API costs depending on your experiment configuration.✓The server runs locally via `uvx` and connects to the LangSmith cloud API using your API key. The server provides read access to your traces, prompts, datasets, and experiments; it can also push prompts (write). Trace data may contain sensitive LLM inputs and outputs from your production systems — treat retrieved traces as confidential.✓All tools are read-only — the server queries trace data but does not modify your application or tracing backend. Trace data may contain sensitive information (request parameters, user IDs, SQL queries) — ensure Claude has appropriate access to this data.
Privacy notes✓LLM traces (including prompts and completions), evaluation annotations, session data, datasets, and experiment results from your Phoenix workspace are surfaced in Claude's context. Your `PHOENIX_API_KEY` is passed as an environment variable — store it in a secrets manager rather than plaintext shell config.✓LLM traces (including prompts and completions), evaluation scores, experiment results, and prompt library contents from your Opik workspace are surfaced in Claude's context. Your `OPIK_API_KEY` is passed as an environment variable — store it in a secrets manager rather than plaintext shell config.✓Conversation threads, prompt content, run inputs/outputs, and dataset examples are surfaced in Claude's context. LANGSMITH_API_KEY is a secret — store it in your environment or MCP client config, never in repositories.✓Distributed trace content including service names, operation names, error messages, HTTP parameters, and LLM prompt/response metadata may be surfaced in Claude's context. For Traceloop backend, `BACKEND_API_KEY` is required and grants access to your Traceloop organization's trace data.
Prerequisites
  • An Arize Phoenix account (cloud at app.phoenix.arize.com) or self-hosted Phoenix instance.
  • A Phoenix API key from your account settings.
  • Node.js 18+ for `npx`.
  • An MCP client such as Claude Code or Claude Desktop.
  • An Opik account (cloud at comet.com or self-hosted) and an API key from your workspace settings.
  • Python with `uv` installed: `pip install uv` or `brew install uv`.
  • An MCP client such as Claude Code or Claude Desktop.
  • A LangSmith API key from smith.langchain.com (free tier available).
  • Python package manager `uv` installed (for `uvx`).
  • An MCP client such as Claude Code or Claude Desktop.
  • An OpenTelemetry-compatible tracing backend: Jaeger (local), Grafana Tempo, or Traceloop Cloud.
  • Services instrumented with OpenTelemetry sending traces to your backend.
  • Python 3.11+ with `uvx` available.
  • An MCP client such as Claude Code or Claude Desktop.
Install
claude mcp add phoenix -e PHOENIX_HOST=https://app.phoenix.arize.com -e PHOENIX_API_KEY=your-key -- npx -y @arizeai/phoenix-mcp@latest
claude mcp add opik -e OPIK_API_KEY=your-api-key -- uvx opik-mcp
claude mcp add langsmith -e LANGSMITH_API_KEY=your-key -- uvx langsmith-mcp-server
claude mcp add opentelemetry -e BACKEND_TYPE=jaeger -e BACKEND_URL=http://localhost:16686 -- uvx opentelemetry-mcp
Config
{
  "mcpServers": {
    "phoenix": {
      "command": "npx",
      "args": ["-y", "@arizeai/phoenix-mcp@latest"],
      "env": {
        "PHOENIX_HOST": "https://app.phoenix.arize.com",
        "PHOENIX_API_KEY": "your-api-key"
      }
    }
  }
}
{
  "mcpServers": {
    "opik": {
      "command": "uvx",
      "args": ["opik-mcp"],
      "env": {
        "OPIK_API_KEY": "your-api-key",
        "OPIK_WORKSPACE": "default"
      }
    }
  }
}
{
  "mcpServers": {
    "langsmith": {
      "command": "uvx",
      "args": ["langsmith-mcp-server"],
      "env": {
        "LANGSMITH_API_KEY": "your-langsmith-api-key"
      }
    }
  }
}
{
  "mcpServers": {
    "opentelemetry": {
      "command": "uvx",
      "args": ["opentelemetry-mcp"],
      "env": {
        "BACKEND_TYPE": "jaeger",
        "BACKEND_URL": "http://localhost:16686",
        "MAX_TRACES_PER_QUERY": "100",
        "BACKEND_TIMEOUT": "30"
      }
    }
  }
}
Citations
ClaimUnclaimedUnclaimedUnclaimedUnclaimed

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.