Phoenix

What is Phoenix?

Phoenix is an open observability platform for LLM apps. Self-hosted Phoenix records traces, runs evaluations, manages datasets and experiments and stores prompts entirely on your infrastructure. Official docs state self-hosting is free with no feature gates; deploy with Docker, Docker Compose, Kubernetes/Helm or one-click cloud templates. The default UI listens on port 6006 with an OTLP collector on 4317.

Where it fits in a 10–200 person business

• Trace and debug an internal RAG chatbot or agent, span by span
• Run offline evaluations and experiments against a prompt/dataset before shipping
• Keep LLM observability data on your own Postgres (or SQLite) instead of a SaaS collector

These use cases are our editorial suggestions, not claims made by the project.

Can a business use Phoenix? (license)

This is a plain-English summary of the license text, not legal advice. Third-party models you run with it have their own licenses.

Data & compliance: does Phoenix phone home?

Checked against the project's own docs and repository on Oct 25, 2026. "Not documented" means we could not find the answer in official sources; it does not mean the feature is missing or that no data is sent. Third-party services you connect (model APIs, vector databases, search engines) have their own data policies. Compare all apps.

What hardware does Phoenix need?

The project does not publish a minimum CPU/RAM figure, so we leave the official value blank rather than guess.

GPU: No. Phoenix is an observability server; GPU is only relevant for models you evaluate elsewhere.

No official min/recommended CPU or RAM figures are published for the Docker or Compose installs. Production docs recommend pinning the image tag and using a persistent volume (Postgres >= 14 officially supported, or SQLite with a mounted disc).

Source: arize.com/docs/phoenix/self-hosting/deployment-options/docker

How much does it cost to host Phoenix per month?

How to deploy Phoenix

Method: Docker / Docker Compose

# simple (ephemeral SQLite inside the container — pin a release tag for production)
docker pull arizephoenix/phoenix
docker run -p 6006:6006 -p 4317:4317 -i -t arizephoenix/phoenix:latest
# UI: http://localhost:6006
#
# Postgres-backed Compose (copy the sample from the docs into docker-compose.yml), then:
# docker compose up --build

From the official instructions: arize.com/docs/phoenix/self-hosting/deployment-options/docker. Always check the source for the current version before you run it.

Known gotchas when self-hosting Phoenix

• Pin arizephoenix/phoenix to a specific version for production (e.g. arizephoenix/phoenix:4.0.0). Floating :latest can change on the next pull. source
• The simple docker run has no persistent volume; recreating the container loses data unless you attach Postgres or a SQLite disc mount. source
• Web UI telemetry is on by default. Set PHOENIX_TELEMETRY_ENABLED=false to disable analytics tracking in the interface. source
• ELv2 forbids offering Phoenix to third parties as a hosted/managed service with a substantial set of its features; internal self-host is fine. source

Alternatives in LLM gateway & observability

• LiteLLM AI gateway that puts 100+ LLM providers behind one OpenAI-compatible API with keys and budgets.
• Langfuse LLM observability with tracing, prompt management, evaluations and cost tracking.
• Opik Open-source LLM observability and evaluation platform for tracing, evals, prompt management and monitoring.

Sources

• GitHub repository: github.com/Arize-ai/phoenix
• Official documentation: arize.com/docs/phoenix/self-hosting
• License text, Gotcha 4: github.com/Arize-ai/phoenix/blob/main/LICENSE
• Hardware requirements, Deploy instructions, Gotcha 1, Gotcha 2: arize.com/docs/phoenix/self-hosting/deployment-options/docker
• Gotcha 3: arize.com/docs/phoenix/self-hosting/security/privacy
• Hetzner Cloud pricing (cost estimate): www.hetzner.com/cloud/regular-performance
• GitHub stars, commits and releases: GitHub REST API, fetched Oct 25, 2026