DASD 4B: benchmark results

Provider: Other.

Aggregate Rating 1504 ± 27, rank #2275 of 10916 rated models, from 28 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Medmarks - MedHallu Hard47.76Score (%)64.8
Medmarks - MedHallu Medium60Score (%)56.8
Medmarks - MedHallu Easy62.85Score (%)46.6
Medmarks - PubMedQA73.27Score (%)38.6
Medmarks - M-ARC30Score (%)36.9
Medmarks - LongHealth Task 184.67Score (%)36.4
Medmarks - SuperGPQA Medicine Hard31.8Score (%)34.1
Medmarks - MedXpertQA Reasoning19.47Score (%)33.5
Medmarks - MetaMedQA60.11Score (%)32.4
Medmarks - SCTpublic59.24Score (%)31.8
Medmarks - LongHealth Task 285.54Score (%)29.5
Medmarks - Med-HALT Reasoning NOTA44.62Score (%)25

Interactive version: theaggregate.ai/model?slug=dasd-4b · How It Works · Data refreshed daily, snapshot 2026-11-08.