Gemini 3 Deep Think gets ‘major upgrade’ aimed at practical applications

Source: 9to5Google

Overview

Deep Think is Gemini’s “specialized reasoning mode,” and Google today announced a major upgrade to let it “solve modern challenges across science, research, and engineering.”

Google worked with scientists and researchers on this update, aiming to use Deep Think to “tackle tough research challenges” that often lack clear guardrails, have no single correct solution, and involve messy or incomplete data. By blending deep scientific knowledge with everyday engineering utility, Deep Think moves beyond abstract theory to drive practical applications.

Benchmark Performance

• Humanity’s Last Exam – new standard of 48.4% (without tools)
• ARC‑AGI‑2 – unprecedented 84.6%, verified by the ARC Prize Foundation
• Codeforces – Elo rating of 3455 on competitive programming challenges
• International Math Olympiad 2025 – gold‑medal level performance

The upgrade also boosts performance in chemistry, physics (including theoretical), and other scientific domains.

Practical Applications

Deep Think enables researchers to interpret complex data and engineers to model physical systems through code. With the updated mode, you can turn a sketch into a 3D‑printable reality: Deep Think analyzes the drawing, models the complex shape, and generates a file ready for 3D printing.

Availability

The Gemini 3 Deep Think upgrade is now available in the Gemini app for Google AI Ultra subscribers. Google is also offering it via the Gemini API—enterprise users can express interest for early access through this form.