AIMarkTechPostabout 1 hour ago
Liquid AI Releases 74M-Parameter Audio Codec
1 min read
Liquid AI releases LiquidCodec, a ~74M-parameter streaming speech tokenizer designed for its LFM3-Audio-2B and LFM3-Omni-2B models. The tool converts speech into tokens for language models and back into audio. Weights are available on Hugging Face, running on a single CPU thread or CUDA GPUs with 24 kHz audio support.
Level
Hype check
Tap to vote and see what everyone thinks.
