OpenAI updates Responses API so agents can build on earlier reasoning with fewer tokens

OpenAIDevs · x · 2026-11-20

OpenAI's developer account announced that agents using GPT-6.2 in the Responses API can now build on earlier reasoning, delivering better results with fewer tokens and more cache hits.

Starting with the newest frontier models, each reasoning item is tied to the conversation history before it — editing earlier messages causes the API to drop all reasoning that follows the change. Developers should keep history unchanged and append new turns instead.

Related event: OpenAI Responses API Lets Agents Reuse Reasoning History to Cut Tokens(3 posts)→

Original post →

More from coding & agent

coding & agent channel →