Development

Gemini 3 Flash matches a year-old frontier ARC-AGI score at 500x lower cost

development2025-12-18

Claim

Gemini 3 Flash achieved the same ARC-AGI-1 score o3 reached a year earlier at more than 500x lower cost, and beat GPT-5.2 on cost by 6x.

Reported metric: >500x cheaper

Reporting date: (not necessarily the event date)

Sources

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

Issue 002 put the deflation at 390x in a year. Six days later the comparison is 500x against a single prior model.

Knowledge graphopen in browser →

About
Reported in
Actor
Evidence for
Supersedes
Referenced by

Cite or correct this record

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