Performance Benchmarks vs Predecessor

Google reports that the new model demonstrates a 25% improvement in logical reasoning and code generation accuracy compared to the Gemini 3 Pro series.

Sources

Open the full topic