OpenAI Agents SDK

The OpenAI Agents SDK is a lightweight, open-source framework for building agentic applications in code. It is the production-ready successor to OpenAI’s earlier experiments, built around a few clean primitives: agents that use tools, handoffs that delegate to another agent, guardrails that validate inputs and outputs, and sessions that manage history.

The framing for this directory’s audience is consistent with the rest of this category: this is developer tooling, not a product a founder uses to get a job done. There is no app and no no-code builder. If you want results in your business without writing code, look to the automation and assistant tools instead. The Agents SDK is for the founder building their own agent product.

For that builder, the appeal is its minimalism plus built-in tracing, so you can see exactly what an agent did on each run, which makes debugging far easier than heavier frameworks. It ships for Python and TypeScript, works best with OpenAI models but supports others, and is free (you pay only for model usage). The honest caveat is that being lightweight means complex orchestration is your job, not the framework’s.