LFM2.5-1.2B (Thinking) — benchmark results
Provider: Liquid AI. Released 2025-06-17. Access: Open.
Unified ELO 1143 ± 20, rank #1742 of 1776 rated models, from 280 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Spanish NLU - ScaLA ES | 22.46 | Linguistic acceptability Score (%) | 61.8 |
| AI Chess Leaderboard (Reasoning) | 721 | Elo | 60.4 |
| EuroEval Italian NLU - ScaLA IT | 22.28 | Linguistic acceptability Score (%) | 56.7 |
| EuroEval Portuguese NLU - ScaLA PT | 15.53 | Linguistic acceptability Score (%) | 56.6 |
| EuroEval Catalan NLU - ScaLA CA | 3.31 | Linguistic acceptability Score (%) | 46.4 |
| AA Humanity's Last Exam | 6.07 | Accuracy (%) | 44.5 |
| EuroEval Italian Knowledge | 36.06 | Knowledge Average Score (%) | 43.7 |
| AA IFBench | 41.84 | Accuracy (%) | 43.4 |
| EuroEval Spanish NLU - Sentiment Headlines ES | 35.9 | Sentiment classification Score (%) | 43 |
| EuroEval Spanish Knowledge | 36.8 | Knowledge Average Score (%) | 42.7 |
| EuroEval French NLU - ScaLA FR | 27.82 | Linguistic acceptability Score (%) | 42.6 |
| EuroEval Albanian NLU - ScaLA SQ | 4.46 | Linguistic acceptability Score (%) | 42.5 |
Interactive version: theaggregate.ai/model?slug=lfm2-5-1-2b-thinking · How the rankings work · Data refreshed daily, snapshot 2026-07-22.