gemini-3.1-flash-lite

A developer's benchmark comparing RAG against long-context prompting on 12 Wikipedia articles about space missions found that long-context answered 96% of questions correctly (23/24) using about 14.5%…

A benchmark of 2,040 CRM agent rollouts found that over half of the small open-weights model gemma-4-E4B-it's 'failures' actually contained the correct answer but were graded zero because the model ne…

A developer added an AI chatbot to Echostats, a Spotify stats dashboard, using Google AI's gemini-3.1-flash-lite model to provide personalized insights into users' music taste. The chatbot faced chall…