benchgap
Calibration

BrowseComp → Agents' Last Exam

Agents' Last Exam is estimated from BrowseComp with a Hill curve fitted on 6 models measured on both: y = 0.1026 + (1.2000 − 0.1026)·x^6.00 / (1.05694^6.00 + x^6.00), R² = 0.50, cross-validated error 11.6 pp. It is used for 30 estimates.

Estimated modelBrowseCompAgents' Last ExamSource
Inkling 2 Small90.2%40.8%estimated ± 11.6 pp, low confidence
GPT-5.5 Pro90.1%40.7%estimated ± 11.6 pp, low confidence
GPT-5.4 Pro89.3%39.5%estimated ± 11.6 pp, low confidence
ERNIE 5.5 Flash Preview87.4%36.8%estimated ± 11.6 pp, low confidence
Ornith-1.5-397B86.6%35.7%estimated ± 11.6 pp, low confidence
dots3-note Preview83.3%31.5%estimated ± 11.6 pp, low confidence
Kimi K2.683.2%31.3%estimated ± 11.6 pp, low confidence
Sage-A1.581.7%29.6%estimated ± 11.6 pp, low confidence
Ling 3.0 Flash Sante80.2%27.8%estimated ± 11.6 pp, low confidence
Nemotron 3.5 Super VL 120B A12B78.0%25.5%estimated ± 11.6 pp, low confidence
Nemotron 3.5 Super VL 120B A12B NVFP477.5%25.0%estimated ± 11.6 pp, low confidence
Inkling-Small77.4%24.9%estimated ± 11.6 pp, low confidence
Inkling77.1%24.6%estimated ± 11.6 pp, low confidence
Beam 1.0 501B-A23B FP876.7%24.2%estimated ± 11.6 pp, low confidence
Step 3.7 Flash75.8%23.4%estimated ± 11.6 pp, low confidence
Beam 1.0 501B-A23B NVFP475.6%23.2%estimated ± 11.6 pp, low confidence
Agents-A175.5%23.1%estimated ± 11.6 pp, low confidence
Ling 3.0 Flash72.2%20.4%estimated ± 11.6 pp, low confidence
Nemotron 3.5 Bolt 74B-A9B BF1670.0%18.8%estimated ± 11.6 pp, low confidence
Qwen3.5 397B69.0%18.1%estimated ± 11.6 pp, low confidence
Ornith-1.5-35B-A3B67.6%17.3%estimated ± 11.6 pp, low confidence
Agents-A1-4B66.8%16.8%estimated ± 11.6 pp, low confidence
Nemotron 3.5 Bolt 74B-A9B66.8%16.8%estimated ± 11.6 pp, low confidence
GPT-5.265.8%16.3%estimated ± 11.6 pp, low confidence
Qwen3.5-122B-A10B63.8%15.3%estimated ± 11.6 pp, low confidence
GLM-562.0%14.6%estimated ± 11.6 pp, low confidence
Qwen3.5-27B61.0%14.2%estimated ± 11.6 pp, low confidence
Qwen3.5-35B-A3B61.0%14.2%estimated ± 11.6 pp, low confidence
Kimi K2.560.6%14.0%estimated ± 11.6 pp, low confidence
Ornith-1.5-9B56.4%12.7%estimated ± 11.6 pp, low confidence

Fitted on

ModelBrowseCompAgents' Last Exam
DeepSeek V4 Flash 073173.2%25.2%
DeepSeek V4 Pro83.4%25.7%
GPT-6 Astra91.5%59.3%
Kimi K3.1 Pro92.1%43.2%
LongCat-2.591.9%38.5%
Step 5 Preview88.7%29.5%