Related Products
What else should we look at alongside OpenAI Agents SDK?
Alternatives
Other frameworks teams evaluate for running agents in their own code.
- LangGraphAgent Framework+25 PositiveLangChain's MIT-licensed runtime for stateful agents built as graphs, with checkpoints and interrupts across model vendors.
- Claude Agent SDKAgent Framework+52 Strong positiveAnthropic's SDK that runs Claude Code's agent loop, file and shell tools, and permission rules inside your service, on Claude models.
- AI SDKAgent Framework+36 PositiveVercel's TypeScript toolkit for model calls, tool loops, and streaming UI across model vendors.
Complements
Where the SDK's model calls are served and routed, and where its traces can go.
- OpenAI APIAI Model API+3 MixedOpenAI's developer platform, which serves the SDK's default Responses API calls and bills its tokens and hosted tools.
- LiteLLMAI Gateway+8 MixedOpen-source gateway that the SDK's beta LiteLLM adapter uses to reach models from other vendors.
- LangfuseAI Observability+46 PositiveOpen-source LLM tracing that can receive the SDK's traces through a custom trace processor.
