LFM2.5-1.2B Instruct: benchmark results

Provider: Liquid AI. Released 2025-06-17. Access: Open.

Unified ELO 1359 ± 1, rank #1250 of 1392 rated models, from 322 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MERA Code - CodeCorrectness84.5Accuracy (%)89.2
AI Chess Leaderboard (Reasoning)755Elo61.3
LIBRA - ruQuALITY71.4Dataset Total Score (%)50
AA IFBench43.81Accuracy (%)48.6
MERA Code - CodeLinterEval48.27pass@1 (%)47.7
EuroEval Catalan NLU - ScaLA CA2.5Linguistic acceptability Score (%)44.5
AA Humanity's Last Exam6.72Accuracy (%)43.8
EuroEval Portuguese NLU - HAREM36.05Named entity recognition Score (%)34.6
EuroEval Spanish NLU - Sentiment Headlines ES32.04Sentiment classification Score (%)34.2
EuroEval Dutch Simplification - Duidelijke Taal44.51Score (%)34.1
LIBRA - ruTPO70.4Dataset Total Score (%)33.3
EuroEval Portuguese NLU - ScaLA PT3.91Linguistic acceptability Score (%)32.6

Interactive version: theaggregate.ai/model?slug=lfm2-5-1-2b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-05.