Llama 3.2 1B: benchmark results

Provider: Meta. Released 2024-09-25. Access: Open.

Unified ELO 1315 ± 1, rank #1331 of 1392 rated models, from 497 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
CLEAR3.2Version (self-reported)100
LA Leaderboard - GalCoLA53.19Accuracy (%)72.1
Open Japanese LLM - Mbpp Pylint Check59.64Score (%)60.6
Open Japanese LLM - CG10.44Score (%)50.8
EuroEval Polish Summarization - PSC21.7Score (%)50.3
MERA - ruHumanEval10.67pass@1 (%)48.1
LA Leaderboard - AQuAS64.79Accuracy (%)45.6
EuroEval Catalan NLU - Guia CAT54.46Sentiment classification Score (%)45
MERA - ruCodeEval7.01pass@1 (%)44.5
EuroEval Spanish Summarization - Mlsum ES23.3Score (%)42.2
EuroEval Bosnian Summarization - LR SUM BS23.96Score (%)41.4
EuroEval German Summarization - Mlsum DE30.99Score (%)41.2

Interactive version: theaggregate.ai/model?slug=llama-3-2-1b · How It Works · Data refreshed daily, snapshot 2026-09-05.