LFM3 by Liquid AI
Product Hunt Rohan Chaubey ● Covered by 3 sources
Liquid AI launched LFM3, a new set of six small models for phones, laptops and robots. They’ve got open weights and one text model says it can fit in about 1 GB.
Based on reporting by Product Hunt, Rohan Chaubey — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Liquid AI has shipped LFM3, the third generation of its Liquid Foundation Models. The first release is a family of six 2B-class models, aimed at phones, laptops, robots, and other edge devices where smaller really does matter.
The lineup is built around a shared backbone, with models for text, vision, audio, omni, and decision tasks. That setup is the point: one core system, multiple modes of use. And Liquid is pushing the open-weights angle hard, with the models posted on Hugging Face.
The most specific number here is the text model. Liquid says the 1.7B version can fit in about 1 GB at 4-bit and supports a 256K context window. That’s a lot of room for a model this size, at least on paper.
Liquid also says these models can run on phones, laptops, and edge hardware, and can be fine-tuned on a single GPU. That puts LFM3 squarely in the small-model race, where practicality matters more than bragging rights and where getting things to run locally is still a big selling point.
My take — AI-written commentary, not fact-checked reporting
Open weights still matter more than most AI companies want to admit. A model that can sit on a phone or laptop without begging for a cloud budget is the kind of unglamorous progress that actually changes how people use AI. The future is not one giant chatbot in a rented data center; it’s smaller models doing useful work where the data already lives.
Read more about this at: Product Hunt
Related stories
Liquid AI Releases LFM2.5-2.6B: An On-Device Agentic Model With 128K Context, Tool Calling, And Open Weights
MarkTechPost · 2 months ago ·
50
Liquid AI Releases LFM2.5-VL-3B: A 3B Vision-Language Model That Reads Screens, Grounds Objects, and Calls Tools On-Device
MarkTechPost · 2 months ago ·
26
Liquid AI Releases LFM2.5-24B-A2B: A 2B-Active MoE for On-Device Agents
MarkTechPost · 2 weeks ago ·
35