Skip to content

Benchmark

Vals MedScribe

Scores are percentages. Higher is better.

Models measured
103
Scores recorded
207
Sources
2
IQ weight
2.20

Score distribution

78.6%
Mean
80.8%
Median
11.3%
Std dev
67.7%
10th percentile
88.5%
90th percentile
4.3%
Min
4.3%91.4%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.20 for chat models, measured over 103 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201Mercury 2.5clinical-result-2240· Clinical Benchmarks data docs55.1%
202Llama 4 Maverickclinical-result-2241· Clinical Benchmarks data docs54.2%
203Llama 4 Maverick· Vals AI54.2%
204Llama 4 Scout· Vals AI50.6%
205Llama 4 Scoutclinical-result-2242· Clinical Benchmarks data docs50.6%
206Nemotron 3.5 Lightningclinical-result-2243· Clinical Benchmarks data docs4.3%
207Nemotron 3.5 Lightning 30B A3B· Vals AI4.3%

Showing 201–207 of 207 results.

Data credits: Clinical Benchmarks (clinicalbenchmarks.ai), CC BY 4.0; fractions converted to percentages; source documents retain their own terms (licence)