Google AI expands Gemini 3 to 120 countries while OpenAI fires back with code red alert

Google AI just made a bold move that’s sending shockwaves through Silicon Valley. Gemini 3, the company’s most intelligent AI model yet, has expanded to 120 countries while delivering breakthrough capabilities in reasoning and multimodal understanding. Meanwhile, OpenAI CEO Sam Altman declared code red over competitive threats, signaling renewed urgency in the AI arms race.

Google’s Gemini 3 Sets New Standards for AI Reasoning

Gemini 3 represents Google’s most comprehensive AI breakthrough to date, combining state-of-the-art reasoning with unprecedented multimodal capabilities. The model tops LMArena Leaderboard with a revolutionary 1501 Elo score, demonstrating superiority across a vast array of benchmarks.

Gemini 3 Pro achieves PhD-level reasoning with top scores on Humanity’s Last Exam at 37.5% and GPQA Diamond at 91.9%. For mathematics, it sets a new frontier benchmark of 23.4% on MathArena Apex. In multimodal testing, the model scores 81% on MMMU-Pro and 87.6% on Video-MMMU, showing dramatic improvement in visual reasoning.

According to Google CEO Sundar Pichai, Gemini 3 combines depth and nuance in ways previous models couldn’t achieve. The model integrates 1 million-token context window with enhanced multimodal capabilities, enabling users to process vast amounts of information across text, images, video, audio, and code simultaneously.

The 120-Country Expansion Brings Gemini Power to Global Markets

As of December 1-2, 2025, Google AI expanded Gemini 3 and Nano Banana Pro to nearly 120 countries and territories in English through AI Mode. This represents the most aggressive international rollout for Google’s AI technology to date, making advanced reasoning accessible to billions of users worldwide.

The expansion covers Americas, Asia-Pacific, and EMEA regions, with support for English queries across all territories. Google AI Mode in Search now features Gemini 3 with dynamic visual layouts and interactive simulations generated on-the-fly based on user queries. Users in supported regions can now access advanced AI assistance while searching, combining Google’s search infrastructure with Gemini 3’s reasoning power.

Feature	Details
Global Availability	120 countries and territories in English
Launch Date	December 1-2, 2025
Supported Platforms	Gemini app, AI Mode in Search, AI Studio, Vertex AI, Google Antigravity
Key Model	Gemini 3 Pro with Deep Think mode coming soon

OpenAI Declares Code Red as Competitive Pressure Mounts

On December 2, 2025, OpenAI CEO Sam Altman sent a companywide “code red” memo signaling urgent action to improve ChatGPT amid intensifying competition from Google AI and other rivals. According to multiple reports from Wall Street Journal, CNBC, and AP News, Altman directed all hands to focus on enhancing ChatGPT’s performance and reliability.

The memo represents the most decisive indication yet of pressure OpenAI faces from competitors. Notably, this echoes a similar situation from December 2022 when Google CEO Sundar Pichai declared code red in response to ChatGPT’s sudden viral success. The irony is striking: OpenAI faces the same existential competitive threat that Google felt nearly three years earlier.

“OpenAI will push back its rollout of other products to focus on improving ChatGPT amid competitive threats from Google and Anthropic.”

— CEO Leadership At OpenAI, Via Yahoo Finance reporting

Breaking Down Gemini 3’s Advanced Reasoning and Capabilities

What makes Gemini 3 revolutionary is not just its benchmark scores, but its fundamental approach to problem-solving. The model features “Deep Think” mode, an enhanced reasoning system that pushes performance boundaries even further for complex challenges requiring intense analysis.

Gemini 3 Deep Think outperforms standard Gemini 3 Pro on Humanity’s Last Exam with 41.0% accuracy and achieves unprecedented 45.1% on ARC-AGI-2, demonstrating its ability to solve novel challenges. The model excels at tool use, achieving 54.2% on Terminal-Bench 2.0 for computer operation via terminal commands.

For developers, Gemini 3 combines exceptional agentic coding with visual generation. It tops WebDev Arena leaderboard at 1487 Elo and scores 76.2% on SWE-bench Verified, making it the best coding agent Google has built. The model can now independently plan and execute complex end-to-end software tasks while validating its own work.

What Does This Mean for the Future of AI Competition?

The convergence of Google’s aggressive Gemini 3 expansion and OpenAI’s code red alert signals a critical inflection point in AI development. Google’s full-stack approach — combining superior hardware infrastructure (TPUs), world-class research, leading-edge models, and access to billions of users through Search and Android — gives the company structural advantages.

For consumers and businesses, this competition accelerates innovation dramatically. 120-country expansion means AI-powered reasoning is becoming global infrastructure rather than specialized technology. Meanwhile, OpenAI’s urgency suggests rapid improvements to ChatGPT coming soon, intensifying the arms race further.

Sources

• Google Official Blog — Gemini 3 launch announcement and technical specifications
• Wall Street Journal — OpenAI code red memo and competitive analysis
• CNBC — OpenAI leadership statements and market implications