OpenAI has introduced WebSocket support to their Responses API, significantly accelerating AI agent operations by reducing waiting times by up to 40% when using external tools. This improvement stems from maintaining a continuous two-way data stream, allowing the server to locally store recent interactions, and thereby minimizing the data sent during sequential tasks. As a result, developers using tools like coding assistants report a 30% increase in responsiveness. This enhancement aligns with the industry trend towards persistent connections in AI APIs, which are becoming essential for more efficient background processing and agent autonomy.
OpenAI adds WebSocket to Responses API for faster AI agents
