LFM2.5-VL-1.6B: benchmark results

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

Unified ELO 1396 ± 1, rank #1153 of 1392 rated models, from 40 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AI for Education Visual Reasoning - match (process)30.6Accuracy (%)34.6
AA Humanity's Last Exam5.1Accuracy (%)32
AI for Education Visual Reasoning - pattern completion (2d)31.5Accuracy (%)26.2
AA CritPt0Accuracy (%)24.4
AA IFBench33.13Accuracy (%)21.7
AA Omniscience - Software Engineering (SWE) - R2Accuracy (%)14
AA Omniscience - Software Engineering (SWE) - Julia0Accuracy (%)10.8
AI for Education Visual Reasoning - reasoning by analogy26Accuracy (%)10.8
HarmVideoBench66.8Macro Avg. (self-reported)10
Artificial Analysis Intelligence Index1Intelligence Index9.4
AA Omniscience - Software Engineering (SWE) - TypeScript5.56Accuracy (%)8.7
AA Omniscience - Software Engineering (SWE) - Rust22Accuracy (%)7.5

Interactive version: theaggregate.ai/model?slug=lfm2-5-vl-1-6b · How It Works · Data refreshed daily, snapshot 2026-09-05.