GitHub radar
Install a lightweight AI model on your computer, no internet needed
LFM3-1.7B is a compact text model from LiquidAI that you can download and run right on your own laptop, with no cloud and no subscription.
It's a small AI model for text that understands sixteen languages and runs entirely on your own device.
When it helps
Useful if you want a local assistant for simple tasks — notes, translation, short text — without internet. For serious coding or reasoning, use a stronger cloud model instead.
Pros
- Runs offline, no subscription, no data sent to the cloud
- Understands sixteen languages, including English and Russian
- Comes in several file sizes — pick one that fits your computer's memory
- The smallest version was specially retrained to lose less quality
Cons
- Weaker than large models at coding and complex reasoning
- You must pick and configure the right file for your computer yourself
How to set it up — step by step
- 1Open the model page on Hugging Face: https://huggingface.co/LiquidAI/LFM3-1.7B-GGUF
- 2Download a GGUF file (a compressed model format) — pick a size that fits your computer's memory
- 3Install LM Studio or llama.cpp — free apps that run the model on your computer
- 4Run the model with: llama-cli -hf LiquidAI/LFM3-1.7B-GGUF:Q4_K_M --temp 0.1 --top-k 50 --repeat-penalty 1.1
- 5For math and science tasks, use different settings instead: --temp 1.0 --top-p 0.95 --top-k 20
- 6If you're working through an AI agent, send it the text in the block below to have it install the model for you
Text for your agent
Copy this and send it to your agent — Claude Code, Codex, any of them:
Set up the LFM3-1.7B model from LiquidAI locally for me: https://huggingface.co/LiquidAI/LFM3-1.7B-GGUF — grab the ready-made compressed (GGUF) file, run it through llama.cpp or LM Studio, and show me how to ask it a question.
A regular laptop — the compressed version of the model takes under 2 GB, the lightest build around 1 GB (this is an estimate based on the model's size).
Open on Hugging Face▌ More finds