Liquid AI releases LFM3 models small enough to run on a phone

Liquid AI shipped six open-weight models on October 27, including a text model that fits in about a gigabyte of memory at four-bit precision.

Liquid AI announced LFM3 on October 27, a new family of six open-weight models built around a shared 2-billion-parameter backbone, according to the company’s own announcement and reporting from MarkTechPost. The lineup spans text, vision, audio, and decision-making, and the whole point is that none of it needs a datacenter. LFM3-1.7B, the text model, runs with a 256,000-token context window and fits in roughly 1GB of memory at four-bit precision, small enough for a phone or a laptop without a dedicated GPU.

The release lands while AI compute remains expensive and constrained, making smaller models that can run locally an alternative to relying entirely on cloud infrastructure. LFM3 is Liquid AI’s bet that specialized, on-device models can provide a way around some of those costs.

The architecture behind it is what Liquid AI calls tri-hybrid: 24 layers split between 13 gated short convolutions, 5 sliding-window attention layers, and 6 global attention layers, found through what the company describes as a hardware-in-the-loop search across 14 device and backend combinations, including a Raspberry Pi 5 and a Qualcomm Hexagon NPU. That’s not a theoretical efficiency claim. It’s an architecture tuned against the actual chips a phone or a laptop ships with, rather than against a GPU cluster that most startups can’t afford to rent anymore.

Small is easy. Competitive is the harder claim, and Liquid AI backs it with numbers. LFM3-1.7B averages 49.37 across 29 benchmarks, ahead of Alibaba’s Qwen3.5-9B at 47.24, a model more than five times its size, according to MarkTechPost’s review of the release. On IFEval, a test of how well a model follows instructions, LFM3-1.7B scores 96.66. On AutomationBench, a tool-use and agentic benchmark, it hits 41.29 against 15.10 for Qwen3.5-9B. The vision model, LFM3-VL-2B, is Liquid’s first to accept video input and averages 71.4 across 62 benchmarks, up from 57.6 for its own predecessor, LFM2.5-VL-3B, despite having fewer parameters.

The audio model, LFM3-Audio-2B, supports 18 languages and posts a speech-quality score of 3.83 on UTMOSv2 against 3.22 for the prior LFM2.5-Audio-1.5B, with a causal encoder that gets time-to-first-token down to as low as 50 milliseconds. There’s also LFM3-Omni-2B, which handles multiple modalities at once, and two decision models aimed at agentic tasks rather than conversation. Six models, one shared backbone, each one specialized rather than general-purpose.

That specialization is the real pitch to builders, not just the file size. Liquid AI says the models can be adapted on a single GPU, which matters enormously for a startup working with limited compute. Weights are downloadable from Hugging Face. Arm was among the launch partners listed by Liquid AI.

None of this makes the demand for cloud compute disappear. Training genuinely large frontier models still requires substantial GPU infrastructure, and nothing about a 2B-parameter phone model changes that math for the labs racing toward bigger systems. What it does is give startups that want an AI feature without a large cloud bill an alternative that can run locally. Liquid AI still competes in a crowded field here: Google’s Gemma, Mistral’s small models, and Alibaba’s Qwen line are all chasing the same efficient, on-device niche, and Qwen3.5-9B is one of the models Liquid AI compares with LFM3 in its benchmark table.

Liquid AI raised $250 million in a Series A in December 2024 at a reported $2.3 billion valuation. The round was led by AMD, with participation from OSS Capital, Duke Capital Partners, PagsGroup and other investors. Back then, on-device efficiency was a compelling architectural story. In October 2026, amid sustained demand for AI compute, it’s starting to look like the plan.

Also read: SpAItial launches Echo-3 to turn iPhone scans into 3D digital twins • ASMPT sales jump 49% on AI chip packaging as core equipment unit misses • Wall Street keeps betting on IREN’s pivot from bitcoin miner to AI cloud

This article is posted in AI News, check it out for more related stories.