Development

Gemini 3.1 Pro takes the index lead at half the evaluation cost

development2026-02-20

Claim

Google launched Gemini 3.1 Pro with 44.4% on Humanity's Last Exam without tools, 77.1% on ARC-AGI-2 and 94.3% on GPQA Diamond, which Artificial Analysis ranked the new Intelligence Index leader at less than half the evaluation cost of Opus 4.6 or GPT-5.2.

Reported metric: 44.4% HLE / 77.1% ARC-AGI-2

Reporting date: 2026-02-20 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to February 20, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

About

Gemini 3.1 Pro

Reported in

Welcome to February 20, 2026

Actor

Evidence for

Supersedes

The cheap model beats the expensive one

Referenced by

Cite or correct this record

Citation formatCitation text

Gemini 3.1 Pro takes the index lead at half the evaluation cost
Innermost Loop KB — secondary record
Reporting date: 2026-02-20
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-02-20-gemini-31-pro-leads-the-index
Claim: Google launched Gemini 3.1 Pro with 44.4% on Humanity's Last Exam without tools, 77.1% on ARC-AGI-2 and 94.3% on GPQA Diamond, which Artificial Analysis ranked the new Intelligence Index leader at less than half the evaluation cost of Opus 4.6 or GPT-5.2.
Attributed sources:
- Welcome to February 20, 2026 — https://theinnermostloop.substack.com/p/welcome-to-february-20-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.