Gemini 3.1 Pro tackles questions that need more work

Google introduced Gemini 3.1 Pro and began rolling it out through the Gemini app, developer tools, and enterprise channels. The announcement focuses on complex reasoning and combining information from several inputs, supported by the company’s evaluations and example tasks.

Context

A useful test is a concrete job with a checkable outcome: reconcile documents, explain an ambiguity, or find an error. Improvements in overall scores need not be uniform across those cases. Comparing the new model against familiar work makes it easier to see where an upgrade actually reduces effort.