LFM3

Liquid AI's new 2B-class family: a 1.7B text model with 256K context that fits in 1 GB, plus vision, audio and omni siblings.

In short. LFM3 is a family of very small AI models from Liquid AI that run on phones and laptops. The main one has 1.7 billion parameters and needs only about 1 GB of memory when compressed. Sibling models add vision, voice and video. They are meant to be fine-tuned for one job rather than used as all-purpose chatbots.

Summary

Liquid AI announced LFM3 on 27 October 2026 as the next generation of its Liquid Foundation Models. The first release has six models of about 2 billion parameters each, all branched from one 1.7-billion-parameter base model trained on 22 trillion tokens: text, vision, audio and omni models, plus two "decision" models specialized from the vision and omni ones. Base and post-trained weights are open on Hugging Face.

The architecture is a "tri-hybrid" of three layer types. LFM3-1.7B has 24 layers: 13 gated short convolutions, 5 sliding-window attention layers and 6 global attention layers. Liquid says it beats its previous LFM2.5-2.6B with 35% fewer parameters.

LFM3-1.7B is a text model post-trained for agent harnesses such as Hermes Agent and OpenClaw, with a 256K-token context and 16 languages. At 4-bit precision it takes about 1 GB. LFM3-VL-2B adds image and, for the first time in an LFM, video input and step-by-step reasoning. LFM3-Omni-2B, with 2.17 billion parameters, takes text, images, video and audio and answers in text or speech; its context is 32K tokens.

Liquid's own benchmark tables place LFM3-1.7B ahead of several larger models on tool use and instruction following. The model card is more cautious: it recommends retrieval or search for knowledge-heavy tasks and says the model is not recommended for agentic coding or complex multi-step reasoning because of its small size. The license is the LFM Open License v1.0, which is free for commercial use only below US$10 million of annual revenue.

Good for

  • On-device agents and tool calling on phones, laptops and small boards
  • Fine-tuning a small model on one domain or file format with a single GPU
  • Voice, vision or video assistants running locally (the VL and Omni variants)
  • Startups and small companies, since commercial use is free below US$10 million of revenue

Watch out for

  • Commercial use by entities at or above US$10 million of annual revenue is not licensed
  • At this size it is weak on knowledge, agentic coding and complex reasoning; pair it with search or retrieval
  • Omni-2B has a 32K context, much shorter than the 256K of the text model
  • Benchmark comparisons come from Liquid's own tables

Performance claims come from the developer and have not been independently verified.

How much memory does it need?

Estimates for the LFM3-1.7B variant. A mixture-of-experts model must hold all of its experts in memory even though only some are used for each token, so total parameters decide the memory and active parameters decide the speed.

Variants Parameters BF16 FP8 / INT8 4-bit (Q4) Fits on
LFM3-1.7B BF16 1.73B 3.98 GB 1.99 GB 1.19 GB Laptop or phone
LFM3-1.7B-Base BF16 1.73B 3.98 GB 1.99 GB 1.19 GB Laptop or phone
LFM3-VL-2B BF16 1.92B 4.42 GB 2.21 GB 1.32 GB Laptop or phone
LFM3-Omni-2B BF16 2.17B 4.99 GB 2.5 GB 1.5 GB Laptop or phone

Estimates: weights plus about 15% overhead. Real usage grows with context length and depends on the software.

VRAM calculator

The license in plain words

An Apache-2.0-style license from Liquid AI with one big catch. Commercial use is licensed only while your company, together with its affiliates, earns under US$10 million a year. Above that line you are not licensed and need your own agreement.

  • Free commercial use only below a revenue cap

More about LFM Open 1.0

See also

Frequently asked questions

What is LFM3?
LFM3 is a family of very small AI models from Liquid AI that run on phones and laptops. The main one has 1.7 billion parameters and needs only about 1 GB of memory when compressed. Sibling models add vision, voice and video. They are meant to be fine-tuned for one job rather than used as all-purpose chatbots.
Can I use LFM3 commercially?
Yes, with conditions. LFM3 is released under the LFM Open 1.0, which allows commercial use but adds extra rules: Free commercial use only below a revenue cap. Read the license text before you ship a product.
Is LFM3 open source?
LFM3 is open-weight, not open source in the strict sense. You can download the model files, but the LFM Open 1.0 adds conditions that the Open Source Initiative definition does not allow.
Can I run LFM3 on my own computer?
Yes. The smallest variant, LFM3-1.7B, needs about 1.19 GB of memory at 4-bit precision, which fits a modern laptop or a mid-range graphics card.
When was LFM3 released?
LFM3 was released by Liquid AI on October 27, 2026.
How long a context window does LFM3 have?
LFM3 supports a context window of up to 256K tokens.

Official links

Sources

Specifications come from these primary sources. Benchmark claims, where mentioned, are reported by the developer.

Release date and facts last verified November 8, 2026.

Related models

Model Released Parameters Context License Smallest variant needs
LFM3-Audio-2B Liquid AI 2B 32K LFM Open 1.0 (Conditional) 4.6 GB Laptop or phone
LFM2.5-24B-A2B Liquid AI 24B 2.45B active 128K LFM Open 1.0 (Conditional) 17 GB Gaming GPU
Gemma 4 Google DeepMind 5.1B – 30.7B 256K Apache 2.0 (Permissive) 3.52 GB Laptop or phone
Raon-1 nano KRAFTON 3.1B 256K Apache 2.0 (Permissive) 1.98 GB Laptop or phone