Liquid AI Debuts LFM2.5-2.6B: A Powerful AI Agent Model Running on Raspberry Pi Without Cloud or GPUs

August 17, 2026

VentureBeat USA

Overview

AI startup Liquid AI debuted LFM2.5-2.6B, a new open-weight language model designed for agentic workloads that can run on local hardware without cloud inference or GPUs. This model enables edge AI applications on devices ranging from smartphones to Raspberry Pi, offering solutions for regulated industries or sensitive information. The launch also includes a strategic partnership with MacPaw to build an on-device AI stack for Mac, leveraging Liquid AI’s foundation models for MacPaw’s macOS assistant.

In Depth

Key Findings

AI startup Liquid AI has unveiled its groundbreaking new language model, “LFM2.5-2.6B.” This model is specifically tailored for agentic workloads and stands out for its ability to run directly on local hardware without reliance on cloud-based inference or high-performance GPUs. This innovation enables advanced AI agent applications on a wide range of edge devices, from smartphones to ultra-small devices like the Raspberry Pi, opening new possibilities for regulated industries and sensitive information processing where data privacy is paramount.

Technical / Clinical Details

The LFM2.5-2.6B model achieves high performance even in environments with limited computational resources and memory bandwidth, thanks to its optimized architecture and efficient inference techniques. This technological advance allows AI agents to operate fully autonomously within a device, without needing to connect to the cloud. This directly translates to reduced network latency, enhanced data security, and lower operational costs. Liquid AI has also announced a strategic partnership with MacPaw to build an on-device AI stack for Mac, leveraging Liquid AI’s foundation models for MacPaw’s macOS assistant. This means Mac users can enjoy advanced AI functionalities without their personal data being transmitted off-device.

Background & Context

Previous AI models, especially large language models (LLMs), have primarily operated in cloud-based data centers due to their high computational requirements. However, this has led to challenges concerning data privacy, latency, and offline usability. Edge AI, the technology of running AI directly on devices, solves these issues, enabling more personalized, secure, and responsive AI applications. There’s a strong demand from regulated industries (e.g., healthcare, finance) and defense sectors to avoid cloud dependence due to data sensitivity, and Liquid AI’s technology directly addresses this need.

Strategic Significance & Outlook

The introduction of Liquid AI’s LFM2.5-2.6B model has the potential to dramatically enhance edge AI capabilities and revolutionize how AI agents are used. As this technology becomes more widespread, more devices will gain intelligent functionalities, allowing users to benefit from AI even in offline or security-critical environments. The partnership with MacPaw is a significant step towards accelerating the adoption of on-device AI in consumer devices, suggesting Liquid AI could establish a new standard for AI agents. In the future, more powerful AI agents are expected to function more autonomously and personally in our daily digital interactions.

Source: https://venturebeat.com/technology/no-cloud-no-gpus-no-problem-liquid-ais-new-model-lfm2-5-2-6b-brings-powerful-ai-agents-to-devices-as-small-as-a-raspberry-pi

Get our weekly technology intelligence — free

Receive an infographic that lets you judge at a glance whether each field’s analysis report is worth reading.

Subscribe Free — Weekly Tech Intelligence

By subscribing, you’ll receive Troy-Technical’s weekly technology intelligence newsletter.