In a bold strategic move to counter intensifying competition, OpenAI is reportedly developing a new large language model (LLM) internally codenamed 'Garlic'. This development coincides with CEO Sam Altman declaring an internal 'Code Red' to urgently improve the company's flagship product, ChatGPT.
The 'Garlic' Gambit: A Specialist AI Challenger
According to reports from The Information, the new 'Garlic' model is being positioned as a high-performance specialist. Internal evaluations suggest it outperforms leading rivals, specifically Google's Gemini 3 and Anthropic's Claude Opus 4.5, in critical areas like coding and complex reasoning tasks.
Mark Chen, OpenAI's Chief Research Officer, reportedly informed colleagues about the model's strong internal performance. This signals a potential strategic pivot for OpenAI from generalist AI models towards specialised applications in high-value industries such as biomedicine and healthcare.
The model is anticipated to be released as GPT-5.2 or GPT-5.5 by early 2026. Chen, who has overseen major projects like DALL·E and Codex, is leading this initiative despite some senior researchers departing for competitors this year.
'Code Red': A Response to Mounting Pressure
The news of 'Garlic' emerges directly from a state of heightened urgency within OpenAI. On Monday, December 1, Sam Altman issued an internal memo declaring a 'Code Red', urging employees to prioritise enhancements to ChatGPT above other projects, including delayed advertising plans.
The memo emphasised redirecting resources to improve ChatGPT's responsiveness and personalisation features. Altman also hinted at a new model release scheduled for the following week, aiming to get ahead of Google's Gemini 3, with a focus on refining model behaviour and image generation.
This sense of crisis is driven by a rapidly shifting competitive landscape. Google's aggressive rollout of Gemini 3 on November 18 and its integration across Google's ecosystem has been notably fast. Similarly, Anthropic's release of Claude Opus 4.5 last month, touted as a top coding model, adds to the pressure.
Traffic Dips and Valuation Concerns
A major catalyst for OpenAI's 'Code Red' appears to be tangible user metrics. Data highlighted by entrepreneur Deedy Das shows that in the two weeks following Gemini 3's launch, ChatGPT's unique daily active users (7-day average) declined by 6%.
This dip is significant as OpenAI's consumer business and its ambitious revenue targets rely on sustained user growth. The company is reportedly targeting $20 billion in annual recurring revenue (ARR) by year-end, a goal underpinning its lofty $500 billion valuation. This user decline marks a potential first major setback for the AI pioneer.
Meanwhile, Google's Gemini app has seen substantial growth, reaching 650 million monthly active users in October, up from 450 million in July. The competitive irony is stark: three years ago, Google itself declared a 'Code Red' in response to ChatGPT's explosive popularity.
As the AI arms race accelerates, OpenAI's dual strategy of an emergency push to fortify ChatGPT and the longer-term development of a specialised 'Garlic' model underscores the fierce battle for dominance in the generative AI space.
