Mentions of GPT-5.1 Thinking have appeared on the ChatGPT website, indicating preparation for an update from OpenAI. This model is mentioned in internal components as responsible for complex logical processes, likely aimed at multi-step reasoning or performing tasks similar to an agent’s work. Alongside this, there are rumors of an entire GPT-5.1 lineup, which may include Mini, Thinking, and a possible programming version based on Codex, each designed for different user needs and computational resource constraints.
Some of these variants are already being tested in internal environments, sometimes under hidden names, indicating A/B performance testing before public launch. According to reports, GPT-5.1 Thinking may receive a larger context window and reduced API usage costs, affecting both regular users and corporate teams.
The new GPT-5.1 models are expected to have different “thinking budgets,” i.e., levels of available resources for task processing, allowing users to choose the optimal option for their needs. For developers and analysts, this opens up the possibility of using AI for more reliable and consistent results in complex scenarios.
The leak occurred ahead of the anticipated launch of Google’s Gemini 3 Pro, and this may be part of OpenAI’s strategy to respond promptly to new competitive products. There are no official statements regarding the capabilities and release date of GPT-5.1 yet, but the community is closely monitoring for additional information.
