OpenAI Agents SDK
Lightweight Python SDK for building production-grade AI agents
View on GitHubWhat it solves
The bootstrapping problem: going from a one-shot prompt to a real agent loop with handoffs, tool execution, and tracing without writing infrastructure from scratch.
Overview
OpenAI's official Python SDK for building agentic applications. Its core abstraction is the Agent object, a model with instructions, tools, and optional handoff targets. The runtime handles the loop: call the model, execute tools, hand off to sub-agents, loop until done. Replaced the experimental Swarm library in March 2025.
Key facts
- Language
- Python
- License
- MIT
- Maturity
- Community canon
- Maintainer
- openai
- Reviewed
- 2026-06-11
Where to start
Start here to see how an Agent, a Runner, and a tool hook together in about 20 lines.
examples/basic/hello_world_agent.pyTechnologies
More in this domain
LangGraph
Stateful multi-agent orchestration for LLM applications
The state management problem in multi-agent systems: how you build agents that pass context across turns without losing the execution thread.
AutoGen
Multi-agent conversation framework from Microsoft Research
How to coordinate multiple specialized agents toward a shared goal without writing low-level orchestration logic by hand.
CrewAI
Role-based multi-agent orchestration for collaborative task execution
The boilerplate problem in multi-agent setup: defining roles, delegation rules, and task routing without writing a custom orchestration graph.