Benchmark explainer
What is FrontierMath Tier 4?
Research-level mathematical reasoning on novel problems at the frontier of human mathematical expertise.
FrontierMath-Tier-4-v2-Private as reported in Epoch AI's Capabilities Index CSV.
How it's scored
- Metric
- normalized performance in Epoch CSV, rendered as accuracy (%)
- Score ceiling
- 100
- Construction
- 43 private problems (post-v2 correction: 12 corrected, 7 removed); vetted by expert mathematicians across number theory, real analysis, algebraic geometry, category theory.
- Human baseline
- Problems intended to take human researchers hours-to-months; no aggregate published baseline (each problem is its own challenge).
How trustworthy
Discrimination
100/100Does it still separate models?
Saturation headroom
98.5/100How far from ceiling / clustered at the top?
Contamination resistance
100/100Public vs held-out; training-leak risk.
Harness comparability
100/100Is it apples-to-apples, or mixed / vendor-optimized?
Freshness
100/100How old is the benchmark?
Contamination history: Tier-4 problems kept private since release; v2 update corrected 12 problems due to math/typo errors in original set.
Limitations: OpenAI holds exclusive access to a subset (per Epoch's funding); private problems cannot be redistributed for external audit; small set limits granularity.
Who leads FrontierMath Tier 4
| Model | Score | Evidence |
|---|---|---|
| GPT-6.1 Sol | 100 | unverified |
| GPT-6 Astra | 97.6 | unverified |
| Claude Opus 5.5 | 95 | unverified |
| Claude Fable 5 | 90.2 | unverified |
| GPT-6 Sol | 90 | unverified |
Compare the top FrontierMath Tier 4 scorers
Frequently asked questions
FrontierMath-Tier-4-v2-Private: FrontierMath-Tier-4-v2-Private as reported in Epoch AI's Capabilities Index CSV.
GPT-6.1 Sol leads FrontierMath-Tier-4-v2-Private at 100. The full leaderboard above lists every recorded measurement, not just the headline number.
65 models have recorded scores on FrontierMath-Tier-4-v2-Private, spanning a score spread of 100.
tensor.news grades FrontierMath-Tier-4-v2-Private A for integrity (score 100/100), ranking #2 of 65 benchmarks we assess across discrimination, saturation, contamination resistance, harness comparability, and freshness.
No — FrontierMath-Tier-4-v2-Private still has headroom and continues to discriminate between models rather than bunching them at the ceiling.
Its test set is held-out, which lowers — but does not eliminate — the risk that training data leaked into the questions.
Scores are reported under a consistent harness, so comparisons on FrontierMath-Tier-4-v2-Private are reasonably apples-to-apples.
Every FrontierMath-Tier-4-v2-Private measurement is source-backed and tagged reproduced, self-reported, or unverified; the underlying record cites epoch.ai/benchmarks/frontiermath-tier-4-v2.