Black Friday 23 Nov: Insider US$59 a yearRemind me
Back Back
Models6/day by emailPrimary

Epoch AI audited its FrontierMath benchmark after OpenAI flagged suspiciously high scores, and its 12 June 2026 v2 release corrected errors in 42% of the original problems, leaving a verified set of 338 problems across four tiers

Epoch AI, which developed the FrontierMath benchmark, released a major update on June 12, 2026. This update addressed errors in 42% of the original problems. The full FrontierMath dataset now consists of 338 problems, split into Tiers 1-3 (295 problems) and Tier 4 (43 problems).

SourceEpoch AI·June 2026 (v2 release)

Verified 2026-08-24

Track with Insider

Why it matters

Benchmarks can contain errors too, so a score is only as good as the test behind it.

The AI boom,with receipts.

6 sourced facts every weekday, plus the Weekly Brief, AI company deep dives and so much more. Free.

Free · No card · Unsubscribe anytime · See today's issue →

Join now and you start with 100 Receipts on The Receipts Ledger — your first step toward a free year of Insider or Pro. How it works →