Liquid AI Launches On-Device Agentic Model LFM2.5-2.6B
helloiamleonie · x · 2026-08-05
Liquid AI released LFM2.5-2.6B, a new on-device agentic model designed to run entirely locally on phones, laptops, and robots.
- Key Features: Operates fully on-device ensuring data privacy with essentially zero marginal cost per run. Capable of planning, tool calling, and multi-step tasks.
- Architecture & Specs: Utilizes the LFM2.5 flagship hybrid architecture, pre-trained on 34T tokens with a 128K context length.
- Performance: Can handle 32+ parallel sessions on a single GPU with aggregate throughput of hundreds of tokens/sec. Achieves scores comparable to or better than models nearly 4x its size (e.g., Qwen3.5-9B) in various benchmarks.
Related event: Liquid AI Releases 2.6B On-Device Agent Model LFM2.5(18 posts)→
More from Embodied
- Peking University's SimForcing distills simulation motion priors into robot world models — PekingUniversity · 2026-10-16
- NVIDIA and Mercedes-Benz Expand Partnership on Next-Gen Automated Driving — nvidia · 2026-10-16
- ~25,000 Humanoid Robots Shipped in H1 2026, Over 95% from Chinese Vendors — ingliguori · 2026-10-16
- openpilot brings stoplight-stopping autonomy to a 2019 Corolla — k7agar · 2026-10-16
- Hironobu Fujiyoshi publishes deck: from AI agents to Agentic Robotics — HirokatuKataoka · 2026-10-16
- CoDA co-evolves simulation and robot policies via shared failure memory, lifting OOD success from 5.3% to 27.6% — PLAN-Lab · 2026-10-16