Skip to content
Intelligibberish
  • News
  • Articles
  • Guides
  • Tools
  • About

Tag

#ai-race

← All articles

Analysis Feb 20, 2026

Google's Gemini 3.1 Pro Doubles Reasoning Performance, Retakes Benchmark Crown

Google's latest model scores 77% on ARC-AGI-2, more than double its predecessor. At $2 per million tokens, it undercuts competitors while outperforming them on most tests.

Intelligibberish

Independent analysis and commentary on artificial intelligence.

News Articles Guides Tools About Disclosure Privacy RSS

© 2026 Intelligibberish. Signal, not noise.