The 230-million-parameter model runs on smartphones and Raspberry Pis while beating models four times its size on key benchmarks
A model with 230 million parameters just embarrassed competitors packing four times the weight. Liquid AI, the MIT spinout valued at roughly $2 billion, released its LFM2.5-230M foundation model on June 25, targeting on-device AI workflows where cloud access is either impractical or unwanted.
The headline number: LFM2.5-230M scored 22.51 on data extraction tasks using the CaseReportBench dataset. Alibaba’s Qwen3.5-0.8B, a model with 800 million parameters, managed 13.83. Google’s Gemma 3 1B, sitting at a full billion parameters, scraped together 2.28.
Small model, big implications for edge computing #
The LFM2.5-230M was pre-trained on 19 trillion tokens and designed to run, as Liquid AI puts it, nearly “anywhere.” On a Samsung Galaxy S25 Ultra, the model clocks 213 tokens per second. On a Raspberry Pi 5, a single-board computer that costs less than a nice dinner, it still manages 42 tokens per second.
The pre-trained weights are already available on Hugging Face, meaning developers can start building with it immediately.
The model is explicitly designed for data extraction and tool usage in agentic workflows. It won’t write your novel, but it can pull structured data from documents on a phone without ever touching the internet.
Liquid AI’s rapid ascent #
Liquid AI was founded in December 2023 by Ramin Hasani, Mathias Lechner, Alexander Amini, and Daniela Rus, all with roots in MIT’s Computer Science and Artificial Intelligence Laboratory. The company’s foundational research centers on liquid neural networks, a class of architectures inspired by biological neural systems that prioritize efficiency and adaptability.
The funding trajectory has been aggressive. The team raised $37.5 million in seed funding shortly after incorporation. By December 2024, they closed a $250 million Series A led by AMD Ventures, pushing total funding to approximately $293 million and the company’s valuation to around $2 billion.
What this means for investors and the broader AI market #
For the crypto and Web3 ecosystem specifically, Liquid AI has no direct blockchain connections. The company isn’t tokenized, doesn’t operate on any decentralized network, and hasn’t signaled interest in that direction. The risk for investors watching this space is assuming that benchmark performance on a single dataset generalizes broadly. LFM2.5-230M’s dominance on CaseReportBench is impressive, but data extraction is a specific task. Whether Liquid’s architectural advantages hold across reasoning, coding, and general knowledge tasks remains an open question that future releases will need to answer.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our