---
format: "aidr-story-markdown/v1"
id: "9bd0d0a9f7f2c19e57f03e3451af96e7507b5960b199817484a6f1556dcc5c2e"
canonical_url: "https://aidr.today/9bd0d0a9?lang=en"
title: "Liquid AI Launches 74M Parameter Audio Codec for LFM3 Models"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-11-12T22:14:40.000Z"
category: "Frameworks"
topics: ["llm","framework","inference"]
source_urls: ["https://marketbrief.now/ai/liquid-ai-launches-74m-parameter-audio-codec-for-lfm3-models-12cdc8ce","https://huggingnews.com/ai/liquid-ai-launches-74m-parameter-audio-codec-for-lfm3-models-12cdc8ce","https://www.marktechpost.com/2026/11/12/liquid-ai-liquidcodec-streaming-audio-codec/"]
summary: "Liquid AI launched LiquidCodec, a 74M parameter audio codec for LFM3 Models. This streaming audio codec supports LFM3-Audio-2B and LFM3-Omni-2B models. The codec processed speech at 361x real time for encoding and 87x real time for decoding using FP32 precision on a single Apple M4 Max CPU thread."
---

# Liquid AI Launches 74M Parameter Audio Codec for LFM3 Models

> [Open the canonical story](<https://aidr.today/9bd0d0a9?lang=en>)

**Published:** 2026-11-12T22:14:40.000Z
**Category:** Frameworks
**Topics:** llm, framework, inference

## Summary

Liquid AI launched LiquidCodec, a 74M parameter audio codec for LFM3 Models\. This streaming audio codec supports LFM3\-Audio\-2B and LFM3\-Omni\-2B models\. The codec processed speech at 361x real time for encoding and 87x real time for decoding using FP32 precision on a single Apple M4 Max CPU thread\.

## Sources

- [Story source](<https://marketbrief.now/ai/liquid-ai-launches-74m-parameter-audio-codec-for-lfm3-models-12cdc8ce>)
- [Story source](<https://huggingnews.com/ai/liquid-ai-launches-74m-parameter-audio-codec-for-lfm3-models-12cdc8ce>)
- [Story source](<https://www.marktechpost.com/2026/11/12/liquid-ai-liquidcodec-streaming-audio-codec/>)