# Liquid AI Builds On-Device Context Layer for Personal AI Agents

By Simon Yoon

Canonical URL: https://www.tokenpost.com/news/technology/28337
Published: 2026-10-08T17:08:06.000Z
Updated: 2026-10-08T17:08:06.000Z
Section: Technology

> Liquid Context is designed for phones, wearables, PCs and vehicles, with a Mercedes-Benz deployment targeted as early as the second half of 2026.

Liquid AI is developing an on-device context layer for personal AI agents, allowing phones, vehicles and other hardware to maintain a local understanding of users while operating within fixed power and computing limits.

Liquid AI announced Liquid Context Sept. 23 at Snapdragon Summit 2026. The system was optimized for Qualcomm Snapdragon processors and the Hexagon neural processing unit, with support for phones, wearables, PCs and vehicles.

Liquid Context is designed to learn routines, preferences and needs from permitted device signals, maintain that context locally and provide relevant information to agents authorized by the user. Those agents may run on the device, in the cloud or across both environments.

Liquid AI separately optimized the Liquid Context memory layer and its LFM2.5-2.6B agentic model for Snapdragon processors. The approach separates the user’s personal context from the agent that uses it to reason, suggest actions or perform tasks with permission.

“Personal AI starts with understanding how you live and what you need, when you need it,” Liquid AI CEO and co-founder Ramin Hasani said.

Liquid AI is also pursuing an automotive deployment with Mercedes-Benz through a multi-year partnership announced April 23. The collaboration covers Mercedes-Benz models equipped with third- and fourth-generation MBUX in North America and focuses on on-device speech, language understanding and reasoning.

The companies targeted an initial production deployment as early as the second half of 2026. The partnership is intended to support private, local and low-latency AI experiences in selected vehicles.

“With this strategic collaboration, we’re scaling embedded intelligence in the vehicle and improving AI speech capabilities for our customers in North America,” Jörg Burzer, Mercedes-Benz Group AG board member and chief technology officer for development and procurement, said.

Liquid AI is developing observability and continuous-improvement loops intended to improve its models and agent software through natural usage.

“From here, we want to build these systems and these agents to be able to self-heal and improve and personalize on their own over time,” Liquid AI COO Jeffrey Li said.