plamo-2-8B: benchmark results

Provider: Other. Access: Open.

Unified ELO 1507 ± 45, rank #1049 of 2656 rated models, from 26 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
pfgen-bench - Completion Mode - Truthfulness0.94Truthfulness Score98.7
pfgen-bench - Completion Mode - Fluency0.86Fluency Score97.1
pfgen-bench - Completion Mode - Score0.75pfgen Score (mean of three)96.5
pfgen-bench - Completion Mode - Helpfulness0.46Helpfulness Score94.1
Swallow - Pre-trained Japanese - NIILC65.5Character F1 (%)85.7
Swallow - Pre-trained Japanese - WMT20 En-Ja28BLEU71.4
Swallow - Pre-trained Japanese - JSQuAD91Character F1 (%)66.3
Swallow - Pre-trained Japanese - JCommonsenseQA90.9Accuracy (%)61.2
Swallow - Pre-trained Japanese - MGSM50.8Exact Match (%)61.2
Swallow - Pre-trained English - TriviaQA58.4Exact Match (%)55.1
Swallow - Pre-trained Japanese - Average48.1Average Score (%)55.1
Swallow - Pre-trained Japanese - JMMLU53.6Accuracy (%)55.1

Interactive version: theaggregate.ai/model?slug=plamo-2-8b · How It Works · Data refreshed daily, snapshot 2026-09-19.