GPT-5.5 arrived just over a month after GPT-5.4. At a press conference, OpenAI's chief scientist Jakub Pachocki was asked whether the release cadence would accelerate further. His answer: a clear "yes." "We'll see quite a lot of progress in the short term, and extremely significant progress in the medium term," he said. "I think progress over the past few years has been surprisingly slow."
OpenAI co-founder and president Greg Brockman said GPT-5.5 "thinks faster and more efficiently with fewer tokens compared to 5.4, so consumers and businesses can use more cutting-edge AI." Chief research officer Mark Chen predicted it will take about two more years before AI can independently complete an entire end-to-end AI research workflow.
Sam Altman posted a thread laying out three positions: gradual deployment is core to safety strategy, GPT-5.5 is already a "smart model" but will iterate quickly; OpenAI aims to build the most efficient models and inference architectures so everyone gets access to the best technology; and the goal is to become "a platform for all companies, scientists, entrepreneurs, and individuals."