Skip to content

LFMs

Generative models for many kinds of sequence data

Visit official site

What LFMs is

The models are described as a family of generative AI models built for any kind of sequence data, including video, audio, text, time series and signals. They use computing units from dynamical systems, signal processing and numerical linear algebra.

What you can do with it

  • Process long sequential multimodal data
  • Work with 1B, 3B or 40B model sizes
  • Use an optimised 32k token context length

Application scenarios

  • Assess risk and forecasts in financial services
  • Support drug discovery and gene sequence analysis
  • Run smart assistants in consumer devices

Pros & cons

✓ What we like

  • A smaller memory footprint on long inputs
  • One family covers several sequence data types

! What to watch out for

  • The named sizes are 1B, 3B and 40B
  • Access starts from the Liquid Playground or Lambda interface

FAQ

What sizes are listed?

The feature list names 1B, 3B and 40B models.

What data can it handle?

The material lists video, audio, text, time series and signals.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review