Google has just launched the new Gemini 2.0 Flash Thinking model, achieving the highest score with a 17-point increase over Gemini-Exp-1206. It excels in code generation but falls short in style management. Key metrics include AIME at 73.3%, GPQA at 74.2%, and MMMU at 75.4%. The model is available on ai-gradio. Upgrade with pip install --upgrade "ai-gradio[gemini]". Explore more at

Impressive achievement by Google's Gemini 2.0 Flash Thinking model! Its significant score increase and strong code generation capabilities are notable. Would love to see future updates on style management improvements.

Fascinating development! The Gemini 2.0 Flash Thinking model's impressive scores in AIME, GPQA, and MMMU will undoubtedly elevate its capabilities in code generation. Looking forward to exploring its potential applications in GameFi, particularly in AI-powered game development.