ChatGPT's Big GPT-5.2 Upgrade Expected This Week

OpenAI is reportedly operating under a "code red" to deploy improvements to ChatGPT so the chatbot can better compete with Google's Gemini 3 model released last month. A few reports early last week said that Sam Altman issued a code red memo to employees, urging them to prioritize ChatGPT development over other projects. The reports said OpenAI may release a new reasoning model as soon as this week. On Friday, The Verge said the big ChatGPT upgrade expected this week is GPT-5.2, with OpenAI earmarking Tuesday (December 9) as a potential release date for the new model. Apparently, GPT-5.2 is ready to be released this week as a first response to Gemini 3. However, the release date isn't set in stone, and OpenAI may postpone the launch.

Separately, The Information reported last week on a "Garlic" codename for a presumably imminent ChatGPT update. Garlic represents a new architecture for ChatGPT and is expected to compete directly against Gemini 3 and Claude Opus 4.5. The report noted that Garlic may be released officially as GPT-5.2 or GPT-5.5, with OpenAI targeting a release by early 2026.

A few weeks ago, Google's Gemini 3 stunned the industry with its massive performance gains. The new model outperformed competitors in several key AI benchmarks. Then, Anthropic released Claude Opus 4.5, which also topped benchmarks. OpenAI released the GPT-5 models in August and then updated ChatGPT to GPT-5.1 a few weeks before the Gemini 3 and Opus 4.5 releases. Altman's code red memo last week suggests that the competition's new frontier models made an impression on OpenAI leadership.

The Verge didn't associate Garlic with GPT-5.2, but said that OpenAI wanted to release the GPT-5.2 upgrade later in December. Pressure from competitors reportedly convinced OpenAI to bring the release forward to this week. If GPT-5.2 is a first response to Gemini 3, as the blog claims, it may not necessarily be the bigger Garlic upgrade that The Information mentioned. It's unclear what new features GPT-5.2 will offer, with The Verge detailing performance improvements that should help OpenAI catch up to Google. OpenAI is focusing on "speed, reliability, and customizability" instead of "flashy new features."

The Wall Street Journal's report detailing Altman's code red memo mentioned the same ChatGPT improvements before The Verge reported a release date for GPT-5.2. The report did not mention the GPT-5.2 name, but said that a new ChatGPT reasoning model would be released this week, ready to outperform Gemini 3 in various tasks.

As for Garlic, the unreleased ChatGPT model supposedly delivers a key innovation from OpenAI. Apparently, OpenAI has managed to improve pretraining efficiency, creating a smaller model that has the same amount of knowledge as a much larger system. This may improve the chatbot's speed and reduce computing costs. Garlic reportedly shows strong performance in benchmark tests, including programming tasks.