CodeEngine: benchmark results

Provider: Undi95. Access: Open.

Aggregate Rating 1332 ± 15, rank #8131 of 10916 rated models, from 68 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MMLU-by-task - College Computer Science53Accuracy (%)92.5
MMLU-by-task - High School Chemistry46.8Accuracy (%)89.7
MMLU-by-task - Electrical Engineering53.1Accuracy (%)85.4
MMLU-by-task - Abstract Algebra34Accuracy (%)82
MMLU-by-task - Business Ethics57Accuracy (%)81.2
MMLU-by-task - Prehistory64.81Accuracy (%)80.9
MMLU-by-task - Logical Fallacies68.1Accuracy (%)80
MMLU-by-task - Anatomy50.37Accuracy (%)79.6
MMLU-by-task - Computer Security69Accuracy (%)78.8
MMLU-by-task - International Law73.55Accuracy (%)78.8
MMLU-by-task - High School Physics34.44Accuracy (%)78.7
MMLU-by-task - Sociology74.13Accuracy (%)78.2

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