OpenAI moving the responses api from http to websockets such a huge win...

Thank you Blacksmith for sponsoring! Check them out at: https://soydev.link/blacksmith

SOURCES
https://x.com/OpenAIDevs/status/2026025368650690932
https://developers.openai.com/api/docs/guides/websocket-mode

Want to sponsor a video? Learn more here: https://soydev.link/sponsor-me

Check out my Twitch, Twitter, Discord more at https://t3.gg

S/O @Ph4seon3 for the awesome edit 🙏

T3Dotgg's resource offers insights, tutorials, and resources for gamers and esports enthusiasts. Readers can learn about gaming news, esports events, and professional gaming strategies. With articles, reviews, and interviews, T3Dotgg provides  guidance and expertise for staying informed and competitive in the world of gaming.

Theo - t3․gg

OpenAI has shifted from REST/SSE to WebSockets for their API, a change with major implications for AI agents. The core problem: every tool call in an agentic workflow requires resending the entire conversation history to a stateless API layer, causing massive bandwidth waste. With WebSockets, a persistent connection guarantees all requests in a session hit the same server box, enabling stateful context management. This means only new inputs need to be sent per tool call instead of the full history, yielding 20-40% speed improvements and 90%+ bandwidth reduction for agentic runs with many tool calls. The author also explains why prompt caching doesn't solve this problem (it reduces compute, not data transfer), and notes that OpenAI open-sourced the responses API standard, increasing the chance this WebSocket approach becomes industry-wide.

I can't believe nobody's done this before...