Skip to content
AI News HubLIVE
More
Summary only

Liquid AI Releases LFM3: 6 Open-Weight 2B-Class Multimodal Models

Liquid AI has released LFM3, a family of six open-weight, 2B-class models sharing one backbone and covering text, vision, audio, omni and decision tasks. LFM3-1.7B uses a tri-hybrid architecture with a 256K context window and fits in about 1 GB at 4-bit.

SourceMarkTechPostAuthor: Asif Razzaq
Liquid AI Releases LFM3: 6 Open-Weight 2B-Class Multimodal Models
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions

Summary only

Under the source content permission, this page provides an editorial summary, key points, and a link to the original article.

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • Six models branch from one tri-hybrid backbone, each fine-tunable on a single GPU
  • LFM3-1.7B averages 49.37 across 29 benchmarks, above Qwen3.5-9B at 47.24 on Liquid's own eval
  • It leads on agent tasks such as BFCL V4 and WebArena-Verified, but knowledge and agentic coding remain weak spots
  • Free commercial use is capped at $10M annual revenue

Highlights and analysis are generated automatically and may contain errors. Check the original source.

More recent stories on this topic · Models

Liquid AI Releases LFM3: 6 Open-Weight 2B-Class Multimodal Models | AI News Hub