Google releases Olympiad medal-winning Gemini 2.5 ‘Deep Think’ AI publicly — but there’s a catch…
1 min read
Summary
Google’s AI-based Gemini 2.5 Deep Think model has officially launched, albeit in a lesser form to the medal-winning iteration.
The model was originally hailed as a trailblazer after winning a gold medal at the International Mathematical Olympiad (IMO).
The AI variant was designed to offer deeper reasoning and more complex problem-solving skills and was originally awarded top marks by mathematical competition officials.
Google has made the newest version, which it has dubbed a “bronze” model, available through its Gemini mobile app to users of its AI Ultra plan, which costs $249.99 a month.
Early testers have said the model is able to probe unsolved problems and help generate potential proofs, as well as taking code refinement to a new level.
The AI can also be used for scientific research, algorithm design, and creative work involving voxel art or user interfaces, among other use cases.