openai-agents

OpenAI Agents SDK

openai-agents Permissive license MIT Active ★ 28,617 v0.20.0 released 2026-08-11

Install

openai-agents on PyPI

pip

pip install openai-agents

uv

uv add openai-agents

poetry

poetry add openai-agents

Package facts

License	MIT (permissive)
Python support	supports the current Python release (>=3.10)
Install friction	low — pure-Python wheel
Runtime dependencies	7 — griffelib, mcp, openai, pydantic, requests, typing-extensions, websockets
Maintenance	actively maintained — 2 days since the last release
Last repo commit	2026-08-13
First released	2025-03-04
Popularity	one of the top 1,000 most-downloaded packages on PyPI (30-day window, as of 2026-08-13)
Known vulnerabilities	none known (OSV.dev, checked 2026-08-13)

Evidence: openai_agents-0.20.0-py3-none-any.whl

Intended Audience :: DevelopersLicense :: OSI Approved :: MIT LicenseOperating System :: OS IndependentProgramming Language :: Python :: 3Programming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.13Programming Language :: Python :: 3.14Topic :: Software Development :: Libraries :: Python ModulesTyping :: Typed

About openai-agents

from the package's own PyPI description — quoted content, verbatim

OpenAI Agents SDK PyPI (image)

The OpenAI Agents SDK is a lightweight yet powerful framework for building multi-agent workflows. It is provider-agnostic, supporting the OpenAI Responses and Chat Completions APIs, as well as 100+ other LLMs.

> [!NOTE] > Looking for the JavaScript/TypeScript version? Check out Agents SDK JS/TS.

Core concepts:

Agents: LLMs configured with instructions, tools, guardrails, and handoffs
Sandbox agents: Agents preconfigured to work with a container to perform work over long time horizons.
Realtime agents: Build powerful voice agents with gpt-realtime-2.1 and full agent features
[**Voice...

Read as markdown · JSON record · Source repository · Homepage

AI interpretation — verify before relying

AI-generated interpretation of the package facts above; every digit, version, license, or vulnerability id it cites is grounded in the facts already shown on this page

A framework for building multi-agent AI workflows supporting text, voice, realtime, and sandbox agents with tools, guardrails, and human-in-the-loop capabilities.

Low friction: pure Python wheel with seven well-maintained runtime dependencies. Released 2 days ago with active maintenance and high community adoption (top 1000 PyPI tier).

MIT license (permissive) allows commercial and private use with minimal restrictions—standard open-source terms suitable for most projects.

Usage

pip install openai-agents

from openai_agents import Agent, Runner

agent = Agent(name="Assistant", instructions="You are a helpful assistant")
result = Runner.run_sync(agent, "Write a haiku about recursion.")
print(result.final_output)

Requires Python 3.10 or newer; OPENAI_API_KEY environment variable must be set.

Verdict: A production-ready, actively maintained agent framework with zero known vulnerabilities, low install friction, and MIT licensing. The recent release (2 days ago), active repository status, and comprehensive feature set make it suitable for building sophisticated multi-agent workflows.

Needs verification

Whether optional dependencies (voice, redis, docker extras) introduce additional system-level requirements or compilation steps.
Performance characteristics and latency expectations for realtime agents over WebSocket in production.
Compatibility details with the 100+ other LLMs mentioned beyond OpenAI's own APIs.
Whether the seven runtime dependencies (openai, pydantic, requests, websockets, mcp, griffelib, typing-extensions) are all required for basic usage or if some are optional.

multi-agent ai workflow frameworkopenai agents sdkllm agent orchestrationvoice agent builderai agent tools and guardrailsrealtime agent frameworksandbox agent executionagent handoffs and delegation