Benchmark saturation clock: every release
Benchmark saturation clock. 3 releases since 16 Oct 2026.
Each release is the file as the site published it on that day, and its copy does not change. Cite a release when a reader must be able to open the same rows later.
The current file changelog.json These releases in the API The page that shows it
Every release, newest first
latest 2026-10-30| Release | Data checked | Rows | Size | Against the one before |
|---|---|---|---|---|
2026-10-30.csv 2026-10-30.json sha256 feed8ea9b018c92f91f0ceaa16c6fecec07ad5c09885efd07059358a01a9cf44 | 30 Oct 2026 | 15 | 3.1 KB | Rows: 15 changed |
2026-10-22.csv 2026-10-22.json sha256 64c09ed21e3a204520cd839d53b4dcc6a865382ae396bca34682aa0c3220eda0 | 22 Oct 2026 | 15 | 3.1 KB | Rows: 1 added, 14 changed |
2026-10-16.csv 2026-10-16.json sha256 066f00dafc24dbd6503a1bf676587b44916a5e72e02ec6c8fcdc622fcb74cbea | 11 Oct 2026 | 14 | 2.9 KB | First release |
What changed on 30 Oct 2026
rows matched by benchmark15 changed
- FrontierMath-Tier-4-v2-Private source
- FrontierMath-Tiers-1-3-v2-Private source
- OTIS Mock AIME 2024-2025 source
- GPQA diamond source
- MATH level 5 source
- SWE-Bench verified days_running, source
- FrontierMath-2025-02-28-Private days_running, source
- FrontierMath-Tier-4-2025-07-01-Private days_running, source
- SimpleQA Verified days_running, source
- Chess Puzzles days_running, source
- MirrorCode best_score, best_model, days_running, source
- EBR-bench best_score, best_model, days_running, source
- Mystery Game Puzzles days_running, source
- Furniture Assembly days_running, source
- SpireClimb days_running, source
What changed on 22 Oct 2026
rows matched by benchmark1 added
- SpireClimb
14 changed
- FrontierMath-Tier-4-v2-Private source
- FrontierMath-Tiers-1-3-v2-Private source
- OTIS Mock AIME 2024-2025 source
- GPQA diamond best_score, best_model, source
- MATH level 5 source
- SWE-Bench verified days_running, source
- FrontierMath-2025-02-28-Private days_running, source
- FrontierMath-Tier-4-2025-07-01-Private days_running, source
- SimpleQA Verified best_score, best_model, days_running, source
- Chess Puzzles best_score, best_model, days_running, source
- MirrorCode days_running, source
- EBR-bench days_running, source
- Mystery Game Puzzles days_running, source
- Furniture Assembly days_running, source
Cite a release
2026-10-30Benchmark saturation clock, release 2026-10-30. jesseahawkins.com. https://jesseahawkins.com/data/releases/frontier/clock/2026-10-30.csv
sha256 feed8ea9b018c92f91f0ceaa16c6fecec07ad5c09885efd07059358a01a9cf44The hash lets a reader confirm that a downloaded copy is this release, byte for byte.