Pinned post
We release LFM3 today, starting with a family of 2B-class models
2:35
@mlech26l
Pinned post
Introducing 𝙻𝚒𝚚𝚞𝚒𝚍𝙲𝚘𝚍𝚎𝚌, our ~74M-parameter streaming audio codec powering LFM3-Audio-2B and LFM3-Omni-2B.View quoted post on X
An audio codec turns speech into tokens a language model can predict, then converts those tokens back into audio. We built ours around two requirements: tokens that are easy to predict and fast conversion back into speech.
On a single CPU thread on an Apple M4 Max, LiquidCodec runs at approximately:
> 361× real time for encoding
> 87× real time for decoding
Both measurements use FP32 precision.
1/ 🧵
What does it take to get AI onto half the world’s devices? Our CEO @ramin_m_h and CTO Mathias Lechner (@mlech26l) break down the roadmap:View quoted post on X
→ We design the model and the hardware together.
→ Through recursive self-improvement, each generation of our models helps build the next.
→ Most devices today are under-utilized, despite the compute and memory already inside them. We build models efficient enough to put that hardware to work across billions of devices.