Arize AI is an AI engineering and observability platform for LLM applications, agents, and traditional ML systems. The commercial Arize AX platform (with Generative and ML & CV variants) provides tracing, evaluation, experiments, prompt management, and the Alyx AI engineering agent, built on the OpenInference OpenTelemetry conventions. Phoenix is the open-source counterpart used by tens of thousands of developers for local tracing, evaluation, and prompt iteration. Arize is vendor- and framework-agnostic with 30+ instrumentation providers and an OTLP-native ingestion path.
Arize AX is the commercial AI engineering platform covering tracing, evaluation, experiments, prompt management, annotations, and dashboards for LLM applications and agents. Bui...
Phoenix is Arize's open-source LLM observability platform offering local tracing, evaluation, experiments, and prompt iteration. Distributed as a Python package with a local UI,...
OpenInference is Arize's open-source set of OpenTelemetry conventions and instrumentation libraries for LLM applications, agents, RAG pipelines, and frameworks. Provides Python ...
Alyx is Arize's AI engineering agent that helps developers debug traces, create evaluators, build dashboards, and compare experiments inside the Arize AX platform.
aid: arize-ai
url: https://raw.githubusercontent.com/api-evangelist/arize-ai/refs/heads/main/apis.yml
name: Arize AI
type: Index
image: https://kinlane-productions.s3.amazonaws.com/apis-json/apis-json-logo.jpg
tags:
- LLM Observability
- ML Monitoring
- Open Source
- OpenTelemetry
- Phoenix
- Tracing
- Evaluation
description: Arize AI is an AI engineering and observability platform for LLM applications, agents, and traditional ML
systems. The commercial Arize AX platform (with Generative and ML & CV variants) provides tracing, evaluation, experiments,
prompt management, and the Alyx AI engineering agent, built on the OpenInference OpenTelemetry conventions. Phoenix is
the open-source counterpart used by tens of thousands of developers for local tracing, evaluation, and prompt iteration.
Arize is vendor- and framework-agnostic with 30+ instrumentation providers and an OTLP-native ingestion path.
created: '2026-05-23'
modified: '2026-05-23'
specificationVersion: '0.19'
apis:
- aid: arize-ai:arize-ax
name: Arize AX
tags:
- LLM Observability
- Evaluation
- Tracing
- Enterprise
humanURL: https://arize.com/docs/ax
properties:
- url: https://arize.com/docs/ax
type: Documentation
- url: https://app.arize.com/
type: ApplicationURL
description: Arize AX is the commercial AI engineering platform covering tracing, evaluation, experiments, prompt management,
annotations, and dashboards for LLM applications and agents. Built on OpenInference and OpenTelemetry with 30+ provider
integrations, and available in Generative and ML & CV variants for different workloads.
- aid: arize-ai:phoenix
name: Phoenix
tags:
- Open Source
- LLM Observability
- Tracing
- Evaluation
humanURL: https://phoenix.arize.com/
properties:
- url: https://docs.arize.com/phoenix
type: Documentation
- url: https://github.com/Arize-ai/phoenix
type: SourceCode
- url: https://pypi.org/project/arize-phoenix/
type: SDK
description: Phoenix is Arize's open-source LLM observability platform offering local tracing, evaluation, experiments,
and prompt iteration. Distributed as a Python package with a local UI, deployable in notebooks, containers, or self-hosted
servers, and instrumented through OpenInference OpenTelemetry conventions.
- aid: arize-ai:openinference
name: OpenInference
tags:
- Open Source
- OpenTelemetry
- Tracing
- Instrumentation
humanURL: https://github.com/Arize-ai/openinference
properties:
- url: https://github.com/Arize-ai/openinference
type: SourceCode
- url: https://opentelemetry.io/
type: Specification
description: OpenInference is Arize's open-source set of OpenTelemetry conventions and instrumentation libraries for
LLM applications, agents, RAG pipelines, and frameworks. Provides Python and TypeScript instrumentors for OpenAI, Anthropic,
LangChain, LlamaIndex, CrewAI, and dozens more, emitting OTLP-compatible spans consumable by Phoenix, Arize AX, or
any OTel backend.
- aid: arize-ai:alyx
name: Alyx
tags:
- AI Agent
- Engineering Assistant
- Debugging
- Evaluation
humanURL: https://arize.com/alyx
properties:
- url: https://arize.com/alyx
type: Documentation
description: Alyx is Arize's AI engineering agent that helps developers debug traces, create evaluators, build dashboards,
and compare experiments inside the Arize AX platform.
common:
- type: Website
url: https://arize.com/
- type: Documentation
url: https://arize.com/docs/ax
- type: PhoenixDocumentation
url: https://docs.arize.com/phoenix
- type: Blog
url: https://arize.com/blog/
- type: Pricing
url: https://arize.com/pricing/
- type: Login
url: https://app.arize.com/
- type: GitHubOrganization
url: https://github.com/Arize-ai
- type: GitHubRepository
url: https://github.com/Arize-ai/phoenix
- type: GitHubRepository
url: https://github.com/Arize-ai/openinference
- type: LinkedIn
url: https://www.linkedin.com/company/arizeai/
- type: Community
url: https://arize-ai.slack.com/
- type: Features
data:
- name: LLM Tracing
description: Capture spans for LLM calls, retrieval steps, tool invocations, and agent loops via OpenInference OTel.
- name: LLM Evaluation
description: Run built-in and custom evaluators on production traces, experiments, and datasets.
- name: Experiments
description: Compare prompt and model variants over curated datasets with structured logging.
- name: Prompt Management
description: Playground, hub, builder, and versioning for prompts used across applications.
- name: Annotations
description: Capture human feedback on traces and outputs for evaluator development and dataset curation.
- name: Alyx AI Engineer
description: AI assistant for debugging, evaluator authoring, dashboarding, and experiment comparison.
- name: ML Monitoring
description: Drift, data quality, and performance monitoring for traditional ML and computer vision models.
- name: Phoenix OSS
description: Open-source local tracing and evaluation tool runnable in notebooks or self-hosted.
- type: UseCases
data:
- name: LLM Application Observability
description: Monitor production LLM applications with traces, evaluators, and alerting.
- name: Agent Debugging
description: Inspect multi-step agent runs across tool calls and intermediate reasoning.
- name: RAG Quality Monitoring
description: Evaluate retrieval and generation quality over time in RAG systems.
- name: ML Monitoring
description: Detect drift and degradation in classical ML and CV models.
- name: Local Development
description: Iterate on prompts and evals locally with Phoenix before shipping to Arize AX.
- type: Integrations
data:
- name: OpenAI
description: OpenInference instrumentation for OpenAI Chat Completions, Assistants, and Responses APIs.
- name: Anthropic
description: Instrumentation for Anthropic Claude models.
- name: LangChain
description: Instrumentation and evaluators for LangChain chains and agents.
- name: LangGraph
description: Trace and evaluate LangGraph stateful agents.
- name: LlamaIndex
description: Instrumentation for LlamaIndex RAG pipelines.
- name: CrewAI
description: Trace CrewAI multi-agent crews.
- name: DSPy
description: Trace and evaluate DSPy programs.
- name: Vercel AI SDK
description: Instrumentation for Vercel AI SDK applications.
- name: OpenTelemetry
description: OTLP-native ingestion compatible with any OTel collector or backend.
- name: Bedrock
description: Instrumentation for AWS Bedrock model invocations.
- name: Vertex AI
description: Instrumentation for Google Vertex AI and Gemini.
maintainers:
- FN: Kin Lane
email: [email protected]
