LFM2

by Liquid AI

About the model:

Hybrid model for on-device deployment.

Model type:

Text

Model performance class:

Full

Link to the model:

https://huggingface.co/LiquidAI/LFM2-24B-A2B

Base models:

unknown

End model:

LFM2-24B-A2B

End model license:

LFM Open License v1.0

About the organisation:

AI startup.

Link to the organisation:

https://www.liquid.ai/

Model release date:

February 2026

Availability

Base Model Data

No base model data sources found.

End User Model Data

No end model data sources found.

Base Model Weights

No base model weights published.

End User Model Weights

End model weights published on HuggingFace.

https://huggingface.co/LiquidAI/LFM2-24B-A2B

Training Code

No training code found.

Documentation

Code Documentation

No code, so no documentation.

Hardware Architecture

Hardware architecture outlined in preprint.

https://arxiv.org/pdf/2511.23404

Preprint

Preprint published on arXiv.

https://arxiv.org/pdf/2511.23404

Paper

Paper listed as forthcoming.

Modelcard

Modelcard contains information about hardware architecture and training approach, but none about data sources and downstream risks.

https://huggingface.co/LiquidAI/LFM2-24B-A2B

Datasheet

No datasheet found.

Access

Licenses

LFM Open License v1.0, not an OSI-approved license.

https://huggingface.co/LiquidAI/LFM2-24B-A2B/blob/main/LICENSE

Is this information not up to date?

Contribute here ->

Supported by the Centre for Language Studies and the Dutch Research Council. Website design & development © 2024 by BSTN. This version of the index generated 09 April 2026, website content last updated 11 March 2026.