Edition·2026-10-19 · Mon
Fed from·Epoch AI·HF Open LLM·OpenAlex·OpenAI evals
The benchmark record for AI models.
Every reported score is tagged by how it was checked — independently reproduced, vendor self-reported, contradicted, or unverified — and every benchmark is graded on the four signals that decide whether a number still means something. We grade the evaluations, not just the models.
Latest measurement·2026-10-07·35 of 65 benchmarks graded A
How trustworthy the evaluations are
Benchmark integrity
Trust holds the top-left; the bottom-right is where scores max out and belief erodes. The best-graded benchmarks:
A | Chess Puzzles | consistent harness |
A | FrontierMath-Tier-4-v2-Private | consistent harness |
A | FrontierMath-Tiers-1-3-v2-Private | consistent harness |
A | ARC-AGI-2 | consistent harness |
A | DTBench | consistent harness |
A | FrontierMath-2025-02-28-Private | consistent harness |
Top 10 models by rank
Frontier leaderboard
| # | Model | Lab | Leads | Avg gap | Evidence |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra | OpenAI | 13 | 3.0 | unverified |
| 2 | Claude Fable 5 | Anthropic | 5 | 11.9 | reproduced |
| 3 | Claude Opus 5.5 | Anthropic | 9 | 4.1 | unverified |
| 4 | GPT-5 | OpenAI | 4 | 31.1 | reproduced |
| 5 | GPT-5.5 | OpenAI | 4 | 17.4 | reproduced |
| 6 | GPT-6.1 Sol | OpenAI | 4 | 4.7 | unverified |
| 7 | Gemini 4 Argon | Google DeepMind | 4 | 8.5 | unverified |
| 8 | Claude Opus 4.6 | Anthropic | 3 | 25.6 | reproduced |
| 9 | Gemini 3 Pro | Google DeepMind | 2 | 20.2 | reproduced |
| 10 | Llama 3.1-405B | Meta AI | 2 | 37.3 | reproduced |
Leads = benchmarks topped; avg gap = mean points behind SOTA elsewhere. Evidence tags how each score was checked — a score is f(model, harness, protocol), not f(model).
Answer pages
What is the X benchmark?
Head-to-head
Popular matchups
Follow the record
Every new score, contradiction, and analysis as it lands — plain Atom, open in any feed reader. No account, no inbox.
Section map
Explore the record
- Leaderboard497
ranked by evidence
- Benchmarks65
graded on integrity
- Compare69
head-to-head matchups
- Labs37
developers tracked
- Explainers28
what each benchmark measures
- Papers60
source records
- Methodology
how we rank & grade