Liquid AI's LFM2-Longevity models for aging-data analysis
Melvin Vivas · X post · 2026-09-18 · Open on X
Topics: Industry Trends & Job Market, Fine-tuning & Model Customization · Level: intermediate
Summary
Liquid AI released LFM2-1.2B-Longevity and LFM2-2.6B-Longevity, small models built to interpret structured aging data. They cover several data types: clinical records, DNA methylation, transcriptomics, proteomics and genetics. The release is a good example of small, domain-specialized models.
Key points
- Two sizes were released: LFM2-1.2B-Longevity and LFM2-2.6B-Longevity.
- They are specialized for interpreting structured aging data across different data types.
- Clinical records come from NHANES.
- DNA methylation data comes from GEO.
- Transcriptomics data comes from GTEx.
- Proteomics data comes from Olink; genetic data is also covered.
Resources mentioned
- LFM2-1.2B-Longevity · tool · huggingface.co · free
Liquid AI's 1.2B-parameter model specialized for interpreting structured aging data. - LFM2-2.6B-Longevity · tool · huggingface.co · free
Liquid AI's 2.6B-parameter model specialized for interpreting structured aging data. - Liquid AI (@liquidai) on X · website · x.com · free
Liquid AI's family of efficient foundation models, used here as the example of a model you can run in the browser with ONNX.
Also in: Trying Liquid AI Decision Models (d1, d1 ultra) via API (Melvin Vivas on X · notes), Liquid RIPL: Low-Latency Inference Engine for LFMs (Melvin Vivas on X · notes), ONNX + Transformers.js: Run Models Like Liquid LFM in the Browser (Melvin Vivas on X · notes), Try Liquid AI's open d1-3B decision model in Colab (Melvin Vivas on X · notes) and 30 more - NHANES · dataset · cdc.gov · free · open in a browser to verify
The US National Health and Nutrition Examination Survey, a source of clinical and health records. - GEO (Gene Expression Omnibus) · dataset · ncbi.nlm.nih.gov · free
NCBI's public repository of functional genomics data, including DNA methylation data. - GTEx (Genotype-Tissue Expression) · dataset · gtexportal.org · free
A public resource of gene expression (transcriptomics) data across human tissues. - Olink · dataset · olink.com · paid
Olink proteomics data, which comes from Olink's protein measurement platform.
More in Industry Trends & Job Market
- AuK: Open-Source Model for Generating and Editing Speech
- Qwen3.8-Omni-Flash: Qwen's agentic omni-modal model
- TypeSafe AI's Jev and System One models: the launch article
- Jev by TypeSafe.AI: a parallel "System 1" model for fast structured decisions
- Melvin Vivas reacts to OpenAI's "GPT-6 Astra" announcement
- Persimmon: A Model Built to Simulate Human Conversation