Liquid AI Launches LFM2.5: 2.6B Parameter On-Device Agentic Model
JosephJacks_ · x · 2026-08-04
Liquid AI has released the new LFM2.5-2.6B model. Designed as an on-device agentic model, it can plan, call tools, and work through multi-step tasks entirely locally on phones, laptops, and robots, keeping data on-device with essentially zero marginal cost per run.
Built on the LFM2.5 flagship hybrid architecture and pre-trained on 34T tokens, the model supports a 128K context length. Official benchmarks show highly impressive performance, comparable to or better than models nearly 4x its size, such as Qwen3.5-9B and Gemma-4-E4B-it.
Related event: Liquid AI Releases 2.6B On-Device Agent Model LFM2.5(18 posts)→
More from Infra
- Baidu open-sources FlashMask: up to 1.55x faster than FlashAttention-4 for long-context training — BaiduResearch · 2026-10-16
- GYRO quant of Qwen3.8-Flash-Next tested: near Q3 quality at much smaller size — PlasticRevenue4601 · 2026-10-16
- RL train-sample systems are frontier AI's hardest infra problem — and a trade secret — brianryhuang · 2026-10-16
- Local tuning of GLM-5.3-Flash on 2×GB10 lifts decode from 49 to 58 tok/s — EAccelerate_42 · 2026-10-16
- 3x DGX Spark cluster runs GLM-5.3-Flash locally behind a proxy, 1.75M tokens/day at 119W — Teknium · 2026-10-16
- Optuna creator Akiba: when 2 hyperparameters cost $20M, what's left for BO/HPO — Stefania_druga · 2026-10-16