Skip to content
View original post on X: Quoc Le· 42/100AI score42/100

Gemini 3 released with MathArena Apex at 23.4%, ARC-AGI-2 at 31%

AISummary

Gemini 3 has been released, with MathArena Apex rising from under 2% to 23.4% and ARC-AGI-2 reaching 31%. The post presents these jumps on hard benchmarks as a meaningful step forward in AI reasoning.

Post on XView on X
Quoc Le
@quocleix

Gemini 3 is finally out. 🚀

The numbers on the hardest benchmarks are wild. Seeing MathArena Apex go from <2% to 23.4% and ARC-AGI-2 hit 31% feels like a real turning point for reasoning. We're starting to crack problems that used to look impossible.

Huge congrats to the team.

Source: Quoc Le · x.comPublished · added here

Topics

Google / GeminiReasoningModel releasesEvals & benchmarks