LFM2.5-2.6B
TrialPlatforms
A 2.6B-parameter on-device language model optimized for agentic workflows.
Why it's here
Placed in Trial: 4 article(s) of evidence from 3 source(s), led by product launches, with 3 in the last 30 days. Confidence 60%.
Evidence (4)
- 8Hugging Face Blog·8/4/2026model_releaseHugging Face introduces LFM2.5-2.6B for on-device agents
Hugging Face announced LFM2.5-2.6B, a 2.6B-parameter model designed to run capable agents entirely on-device with tool calling and multi-step workflows. The model is positioned for laptops and phones, emphasizing privacy, low memory use, and strong benchmark performance relative to much larger models.
- 7Hugging Face Blog·7/28/2026model_releaseLiquidAI releases LFM2.5 encoders for fast long-context CPU inference
Hugging Face announced two new encoder models, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, designed for long-context inference with strong accuracy and low latency on CPU. The models support up to 8,192 tokens, are trained with masked language modeling, and are positioned for use cases such as classification, intent routing, safety filtering, and PII detection.
- 5Simon Willison·7/20/2026product_launchNativ lets Mac users run open models locally
Nativ is a curated library for selecting and running open AI models locally on Mac hardware. It recommends partner models based on the user's machine and highlights options from Google, Cohere, and Liquid AI with different sizes and context windows.
- 5Hugging Face Blog·2/20/2026product_launchTrain small AI models for free with Unsloth and Hugging Face Jobs
Hugging Face describes a workflow for fine-tuning the small LFM2.5-1.2B-Instruct model using Unsloth on Hugging Face Jobs, with free credits offered to eligible users. The post highlights faster training, lower VRAM usage, and agent-based setup through tools like Claude Code and Codex.