Google updates ‘Gemini Pro’ to version 3.1: what’s new?
The company's flagship model has received many improvements.
Google has released an update for its flagship Gemini neural model, the 3.1 Pro. The new model emphasized advanced logical thinking. On the ARC-AGI-2 test, which measures a model’s ability to construct and solve logical models, the 3.1 Pro more than doubled the 3 Pro, achieving scores of 77% and 31%.
Google says that Gemini 3.1 Pro is designed for tasks that require complex reasoning, where a simple answer isn’t enough.
The company has also improved animation generation, maintaining high quality at any scale. Responding to the same needs as before, 3.1 Pro creates more complex and detailed 3D models and animations.
Users of paid plans such as AI Pro and AI Ultra can try out the updated model in preview mode: it is available in the Gemini and NotebookLM apps, as well as through the Gemini API in Google AI Studio and Google Antigravity.
Credits:
