Liquid AI's small Longevity models beat GPT-5 and Claude Opus on Cell-published aging benchmarks

helloiamleonie · x · 2026-09-17

Liquid AI, partnering with Insilico Medicine, published a Cell cover paper showing fine-tuned Liquid Foundation Models achieving SOTA on biological longevity tasks, outperforming Gemini-3.1-Pro, GPT-5 and Claude Opus. They released LongevityBench (17 tasks spanning NHANES clinical records, GEO DNA methylation, GTEx transcriptomics, Olink proteomics, OpenGenes genetic evidence) plus two specialized models, LFM2-1.2B-Longevity and LFM2-2.6B-Longevity, supporting binary classification, pairwise comparison, multiclass classification and numeric regression. Models and a notebook are publicly available.

Related event: Liquid AI and Insilico Medicine longevity research lands Cell cover, small models beat GPT-5(2 posts)→

Original post →

More from Models

Models channel →