Open-source AI workstation for multi-model chat, agents, local knowledge bases, and image creation.
Freemium
www.github.com
OpenAI Agents SDK
Open-source Python framework for multi-agent workflows, tool use, sandbox tasks, and voice agents
OpenAI Agents SDK is a provider-agnostic Python framework for building multi-agent workflows with tools, handoffs, guardrails, sessions, and tracing. OpenAI Agents SDK serves developers building text, sandbox, realtime, and voice agents and is available as an MIT-licensed open-source package installable with pip or uv.
OpenAI Agents SDKPreview unavailable
At a glance
- Pricing
- FreeOpen-source SDK under the MIT license, available through GitHub and installable with pip or uv. Model-service pricing is not stated.
- Added
- 28 Oct 2026
About OpenAI Agents SDK
OpenAI Agents SDK is an open-source Python framework for developers building multi-agent workflows. OpenAI Agents SDK is provider-agnostic and supports the OpenAI Responses and Chat Completions APIs, as well as 100+ other LLMs.
OpenAI Agents SDK configures agents with instructions, tools, guardrails, and handoffs. Agents can delegate tasks to other agents and take actions through functions, MCP, and hosted tools. Built-in capabilities include human involvement during agent runs, automatic conversation history management, and tracing for viewing, debugging, and optimizing workflows.
OpenAI Agents SDK also supports sandbox agents that inspect files, run commands, apply patches, and preserve workspace state. Realtime agents support server-side voice and multimodal experiences over WebSocket, while voice pipelines combine speech-to-text, an agent workflow, and text-to-speech.
OpenAI Agents SDK is available under the MIT license through its public GitHub repository and can be installed with pip or uv. The source does not state model-service pricing.
OpenAI Agents SDK suits developers who want programmable agent workflows rather than a ready-made application. Python 3.10 or newer is required. Voice and Redis session support require optional installation groups. The local Unix sandbox client supports macOS and Linux; Windows users need a Docker or hosted sandbox client.
What you can do
- Agent delegation through tools and handoffs
- Function, MCP, and hosted tool support
- Input and output guardrails
- Human involvement across agent runs
- Automatic conversation history management
- Built-in agent tracing
- Sandbox agents for file and command workflows
- Realtime voice agents and speech pipelines
Pros and cons
What works
- MIT-licensed open-source framework
- Provider-agnostic support for OpenAI APIs and 100+ other LLMs
- Built-in tracing, sessions, and guardrails
- Supports text, sandbox, realtime, and voice agent workflows
What to watch
- Requires Python 3.10 or newer and a configured development environment
- Voice and Redis session support require optional installation groups
- Windows sandbox workflows require Docker or a hosted sandbox client
- Pull requests are restricted to repository collaborators
Who OpenAI Agents SDK is for
Python developers building multi-agent workflows and voice applications
- Build workflows that delegate tasks between agents
- Connect agents to functions, MCP, and hosted tools
- Run file inspection, commands, and patches in sandbox workspaces
- Build realtime voice and multimodal experiences
- Debug agent runs with built-in tracing
OpenAI Agents SDK FAQ
What is OpenAI Agents SDK?
OpenAI Agents SDK is an open-source Python framework for building multi-agent workflows. OpenAI Agents SDK supports agents configured with instructions, tools, guardrails, and handoffs, alongside built-in tracing, human involvement, and automatic conversation history management.
Is OpenAI Agents SDK free?
OpenAI Agents SDK is available as an open-source project under the MIT license. OpenAI Agents SDK can be installed with pip or uv, and the supplied repository text does not state pricing for model services.
Which models and APIs does OpenAI Agents SDK support?
OpenAI Agents SDK supports the OpenAI Responses and Chat Completions APIs, as well as 100+ other LLMs. OpenAI Agents SDK is described as provider-agnostic and supports agent actions through functions, MCP, and hosted tools.
Can OpenAI Agents SDK build voice agents?
OpenAI Agents SDK supports realtime voice agents and voice pipelines. OpenAI Agents SDK provides server-side voice and multimodal experiences over WebSocket, plus pipelines combining speech-to-text, an agent workflow, and text-to-speech. Voice pipeline support uses an optional installation group.
What does OpenAI Agents SDK require to run?
OpenAI Agents SDK requires Python 3.10 or newer and installation through pip or uv. OpenAI Agents SDK examples require the OPENAI_API_KEY environment variable. The Unix local sandbox client supports macOS and Linux; Windows users need Docker or a hosted sandbox client.
From the community
No reviews yet
Be the first to review this tool. What you write helps the next person choose.
Keep exploring
Alternatives to OpenAI Agents SDK
An open-source skills framework for coding agents, covering design, planning, testing, debugging, and code review.
Free
Local model gateway for routing coding agents across providers, models, and accounts.
Free
An open-source interface for AI conversations, agents, code execution, and Model Context Protocol connections.
Free