Skip to content
Longterm Wiki has not been actively maintained since May 2026. Some data is still updated occasionally, but please don't rely on this site being current or accurate.
Longterm Wiki
Index
Fact·f_tHAA1W30dw·Fact

xAI — Benchmark Score: 93.3

Verdictconfirmed95%
1 check · 7/20/2026

1 → confirmed

Our claim

entire record
Subject
xAI
Property
Benchmark Score
Value
93.3
As Of
2025
Notes
Grok 3 AIME 2025 benchmark: 93.3% success rate

Source evidence

1 src · 1 check
confirmed95%primaryHaiku 4.5 · 7/20/2026

NoteThe source directly confirms the claim. The xAI announcement states: 'We tested these models on the 2025 American Invitational Mathematics Examination (AIME), which was released just 7 days ago on Feb 12th. With our highest level of test-time compute (cons@64), Grok 3 (Think) achieved 93.3% on this competition.' The benchmark score of 93.3% for Grok 3 on AIME 2025 is explicitly stated in the source text, matching the claim exactly. The date context (February 19, 2025) aligns with the 'as of 2025' qualifier in the claim.

Case № f_tHAA1W30dwFiled 7/20/2026Confidence 95%
Source Check: Benchmark Score | Longterm Wiki