Orchestration · Project
OpenAI Agents SDK
OpenAI Agents SDK coordinates Python agent runs with tools, handoffs and guardrails.
+ FollowBuilder facts
- Runtime. The MIT-licensed Python package requires Python 3.10 or later. Source ↗
- Model access. The SDK supports OpenAI Responses and Chat Completions plus other model providers. Hosted inference requires credentials for the selected provider. Source ↗
- Handoffs. A handoff transfers control to another agent, which receives the conversation history by default. An input filter can change what it sees. Source ↗
- Guardrails. Input, output and tool guardrails can trip and stop a run. Input guardrails apply to the starting agent and output guardrails to the agent returning a final result. Source ↗
- Tracing. Built-in traces cover model generations, tool calls and handoffs. Tracing can be disabled for a single run or globally. Source ↗
Sources checked
Publisher update ·
OpenAI Agents SDK for Python 0.25.0 ↗Removes the beta Runloop sandbox provider and changes Pydantic run checkpoints to preserve default values and in-place edits.
Change timeline
0 storiesNo verified releases in this timeline yet.