Gemini 3.7 Flash hits the market: a week after launch, breaking growth records

Comparative News, AI News, Google CEO Sundar Pichai said that in the first week of Gemini 3.7 Flash launch, it broke the previous Gemini model growth record and became Google's fastest growing Gemini model so far. However, Google did not announce the exact scale of growth and scale of use. This model really hits the cost-effective dessert zone. Artificial Analysis gave it a comprehensive intelligence index of 56. The high inference mode output speed was about 340 token/s, each evaluation task cost about 0.40 US dollars, and it entered the Pareto frontier of intelligence and completion time. The cost of the medium inference model was further reduced to $0.26, entering the Pareto frontier of intelligence and cost. In ARC Prize's independent verification, Gemini 3.7 Flash High Inference Mode received 84.6% in ARC-AGI-2 (Abstract Reasoning Benchmark), and cost only $0.25 per task. Google also quickly introduced 3.7 Flash into its products. Gemini App, Antigravity, and AI Studio are already connected, and Google Search is starting to use this model.