WebSocket Responses API mode is available over a single, long-lived WebSocket connection.

The point is lower end-to-end latency for tool-heavy agent workloads through one long-lived connection.

This sits in WebSocket Mode docs and gives Responses API builders a lower-latency option for tool-heavy agents.

• Jul 06, 2026OpenAI upgraded ChatGPT memory to keep context current and reduce stale or contradictory saved memories.
• Jul 06, 2026OpenAI began rolling out ads for Free and Go users in the UK.
• Jul 06, 2026xAI streaming Speech to Text now supports Smart Turn end-of-turn detection.

This is the world we build in. It moves this fast. We move with it.