Google Gemini introduced Gemini 2.5 Pro and Gemini Flash, with new features including Deep Think (multi-step reasoning mode) and Computer Use (controlled actions on desktop and Drive files). Gemini also rolled out Veo 3, enabling high-quality AI-generated videos from text prompts.
Gemini 3 Pro leads as the best overall chatbot in December 2025 Text Arena benchmarks, while GPT-5.2 excels in coding with 80% on SWE-bench Verified.
xAI released Grok 4.1 in late November 2025, contributing to the rapid AI model race with weekly frontier model iterations.
Anthropic introduced Claude Opus 4.5 and Claude 4.5, topping benchmarks in web development and integrated rapidly into Microsoft Foundry, GitHub Copilot, and other platforms.
OpenAI released GPT-5.2 on December 11, 2025, in a 'Code Red' response to competitors, described as the most capable model for professional knowledge work, with integrations into partners like Notion and Zoom.
Google launched Gemini 3 on December 10, 2025, featuring a 400,000-token context window, 128,000 max output tokens, and Deep Think mode for advanced 10-15 step reasoning chains.
Market share data as of December 3, 2025, shows Claude AI at 3.80% with 14% quarterly user growth, Google Gemini at 13.40% with 12% growth, positioning them as fast-growing generative AI chatbots.
Anthropic released Claude 4.5 (also referenced as Claude Opus 4.5 and Claude 3 in comparisons), excelling in reviewing large codebases, long document analysis, complex reasoning, and academic research, with Microsoft integrating it into Foundry, GitHub Copilot, and Copilot Studio.
OpenAI launched GPT-5.2 on December 11, 2025, as a 'code red' response to competitors, described as the most capable model for professional knowledge work, with rapid integration into partners like Notion, Box, Shopify, Harvey, and Zoom.
Google released Gemini 3, featuring breakthrough 10-15 step reasoning chains without losing coherence and Deep Think mode for enhanced benchmark scores, alongside Antigravity, a Gemini-powered coding interface combining chat, terminal, and browser for agentic development.
YouTube launched a closed beta for Playables Builder, a Gemini-powered tool allowing creators to generate mini-games from text, image, or video prompts.[1]