Qwen2.5-14B-Hyperionv4: benchmark results

Provider: CultriX. Access: Open.

Aggregate Rating 1576 ± 33, rank #1305 of 10916 rated models, from 56 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard - GPQA39.77Score98.9
Open LLM Leaderboard - MuSR48.32Score95.5
Open LLM Leaderboard - MMLU-Pro53.64Score94.1
Open Japanese LLM - Mbpp Pylint Check97.39Score (%)91.8
Open Japanese LLM - CG61.65Score (%)90.5
Open LLM Leaderboard - BBH64.72Score90.4
Open LLM Leaderboard - MATH Level 534.74Score85.1
Open Japanese LLM - Jcommonsenseqa Exact Match95.8Score (%)84.6
Open Japanese LLM - ALT E TO J Comet Wmt2290.95Score (%)84.3
Open Japanese LLM - Jnli Exact Match86.52Score (%)81.7
Open Japanese LLM Leaderboard64.97Average Score (%)81.4
Open Japanese LLM - Commonsensemoralja Exact Match90.63Score (%)80.9

Interactive version: theaggregate.ai/model?slug=qwen2-5-14b-hyperionv4 · How It Works · Data refreshed daily, snapshot 2026-11-08.