A family of small Liquid AI models that run on your device for text, images, and sound.
LFM3 is like a tiny clinic inside your device. It can read, see, and hear. No cloud waiting room needed.
It helps apps handle many kinds of input locally. It suits AI agents and expert tools.
Foundation-model
LFM3 is a multimodal foundation-model family for local use.
Multimodal AI
LFM3 can work with text, images, and sound.
Open weights
Its base and trained models come as open weights.
Local-LLM
Its text models can run locally for agents and expert tools.