OpenAI has introduced a new model, GPT-6.1 Terra. According to the company's announcement, on the Agents' Last Exam benchmark, at the medium reasoning setting, the model scores 15 percentage points higher than the previous version, GPT-5.6 Terra, at roughly a third of the cost.
The announcement says the new model is available in ChatGPT, Codex and the API. OpenAI also states that the model's visual reasoning and computer-use capabilities are significantly stronger.