Key Info

Liquid AI's LFM3 model family is now available on Ollama, designed as small models that run anywhere and can be fine-tuned on a single GPU.

Highlights

  • Covers text, vision, audio, omni, and decision model variants
  • The text model fits in about 1 GB at 4-bit quantization
  • Supports a 256K context window
  • Available to run locally via Ollama