Audio codec for Speech LLMs: parallel tokens and streaming
In a post published on October 17, 2025, Meituan presented an open-source audio codec optimized for Speech LLMs, with parallel semantic and acoustic tokens at 16.7 Hz and low-latency streaming decoding.
Source: LongCat-Audio-Codec para Speech LLMs (github.com). Text prepared with AI from this source.
What happened and what to do
In a post published on October 17, 2025, Meituan released an open-source audio tokenizer and detokenizer optimized for Speech LLMs. The approach encodes semantic and acoustic tokens in parallel at 16.7 Hz, uses a low bit rate, and includes a low-latency streaming decoder.
A company building voice applications can evaluate the codec in a prototype and measure quality, latency, bandwidth use, and compatibility with its models and systems. Based on the results, it can implement integration, monitoring, and operation of the audio pipeline on its own infrastructure.
How the consultancy can help
Wendelmaques can diagnose the use case and voice requirements, scope an evaluation, and implement codec integration, monitoring, and operation on the company's infrastructure.
Next step
Send a short description of your voice application and technical requirements to receive a scoped proposal.
Consulting for your project
Infrastructure review, deployment and ongoing operations, with scope and pricing defined in the proposal.
Quoted per project
Request a proposal