← Agents' Last Exam

Agents' Last Exam: Score history

Benchmark scores get restated when harnesses change or models are re-run. This page keeps the trail instead of quietly overwriting it.

Timeline

No historical snapshots recorded yet — the current scores are shown below.

Current snapshot

#ModelScoreObserved
1Gemini 4 Argon39.5%2026-10-20
2DeepSeek V4.1-Flash 101735.4%2026-10-20
3DeepSeek V4.1-Flash 091031.8%2026-10-19
4MiniMax M3.1 Flash29.6%2026-10-19
5Claude Opus 528.6%2026-10-19
6GLM 5.328.5%2026-10-19
7Kimi K327.6%2026-10-19
8GPT 5.6-Sol26.7%2026-10-19
9DeepSeek V4-Pro 081325.7%2026-10-19
10DeepSeek V4-Flash 073125.2%2026-10-19

Overwritten scores

Nothing has been overwritten yet.