Liquid AI Releases 2.6B Model: 30 tok/s on Phone with 128K Context
BTA_Labs · reddit · 2026-08-05
Liquid AI released LFM2.5-2.6B, a small model designed for local devices. It features 2.69B parameters, 128K context, tool calling, and is post-trained for multi-step agent workflows.
Performance (Vendor Benchmarks):
- 30 tok/s on a phone, 220 tok/s on an M5 Max.
- Under 2.5 GB memory usage; Q4 GGUF is 1.67 GB.
Capabilities:
- Competitive in tool calling, but still lags behind larger models like Qwen3.5-9B in coding and knowledge-heavy tasks.
- The model card advises against using it for agentic coding.
The author suggests this model is ideal as a local worker agent for repetitive tasks like extraction and tool calls, leaving complex planning to larger models.
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