QwQ-32B-AWQ: benchmark results

Provider: Alibaba. Access: Open.

Aggregate Rating 1588 ± 35, rank #1142 of 10916 rated models, from 14 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
EPPCMinerBen (1-Shot) - Subcode Classification45.94Micro-averaged F1 (%) of EPPC subcode classification (choosi91.7
EPPCMinerBen (2-Shot) - Subcode Classification46.98Micro-averaged F1 (%) of EPPC subcode classification (choosi91.7
Phun-Bench - Homophone Recall80.7Accuracy (%) of recovering a four-character chengyu from an 87.5
Phun-Bench - Rhyming Sentence Generation (1-Rhyme)83Exact-match rate (%) of the rhyme group of the generated sen85.7
Phun-Bench - Rhyming Sentence Generation (2-Rhyme)43.7Exact-match rate (%) of the rhyme groups of the generated se85.7
EPPCMinerBen (0-Shot) - Subcode Classification45.35Micro-averaged F1 (%) of EPPC subcode classification (choosi81.8
ReasonScape M12X483.19ReasonScore (hard)69.6
EPPCMinerBen (1-Shot) - Code Classification61.97Micro-averaged F1 (%) of EPPC code classification (sentence-66.7
EPPCMinerBen (2-Shot) - Code Classification63.36Micro-averaged F1 (%) of EPPC code classification (sentence-66.7
Phun-Bench - Contextual Homophone Recognition57.4Recognition rate (%): share of homophonic pun sentences (1,662.5
EPPCMinerBen (0-Shot) - Code Classification60.19Micro-averaged F1 (%) of EPPC code classification (sentence-45.5
EPPCMinerBen (0-Shot) - Evidence Extraction65.11Micro-averaged F1 (%) of EPPC evidence extraction (the minim33.3

Interactive version: theaggregate.ai/model?slug=qwq-32b-awq · How It Works · Data refreshed daily, snapshot 2026-11-08.