# Rethinking Speech Codecs: From Compression to Autoregressive Generative Modeling

arXiv: https://arxiv.org/abs/2609.04237 · HTML: https://mlsanity.com/paper/2609.04237/

## Key facts

- No citing papers in the map yet (data through 19 Oct 2026).
- Placed in Vocoders and speech synthesis, one of 11 directions in Speech and audio language models.
- Builds on 22 of its 37 references that are in the map.
- First posted on arXiv 27 Jul 2026.

## Placed in

- Speech and audio language models / [Vocoders and speech synthesis](https://mlsanity.com/speech/vocoders-and-speech-synthesis/)

## Builds on (most cited first, 22 of 22)

- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://mlsanity.com/paper/2101.00190/) (2021)
- [Taming Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2012.09841/) (2020)
- [HiFi-GAN: Generative Adversarial Networks for Efficient and High Fidelity Speech Synthesis](https://mlsanity.com/paper/2010.05646/) (2020)
- [SoundStream: An End-to-End Neural Audio Codec](https://mlsanity.com/paper/2107.03312/) (2021)
- [Moshi: a speech-text foundation model for real-time dialogue](https://mlsanity.com/paper/2410.00037/) (2024)
- [High-Fidelity Audio Compression with Improved RVQGAN](https://mlsanity.com/paper/2306.06546/) (2023)
- [SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities](https://mlsanity.com/paper/2305.11000/) (2023)
- [UTMOS: UTokyo-SaruLab System for VoiceMOS Challenge 2022](https://mlsanity.com/paper/2204.02152/) (2022)
- [BEATs: Audio Pre-Training with Acoustic Tokenizers](https://mlsanity.com/paper/2212.09058/) (2022)
- [GLM-4-Voice: Towards Intelligent and Human-Like End-to-End Spoken Chatbot](https://mlsanity.com/paper/2412.02612/) (2024)
- [Mini-Omni: Language Models Can Hear, Talk While Thinking in Streaming](https://mlsanity.com/paper/2408.16725/) (2024)
- [UniAudio: An Audio Foundation Model Toward Universal Audio Generation](https://mlsanity.com/paper/2310.00704/) (2023)
- [LauraGPT: Listen, Attend, Understand, and Regenerate Audio with GPT](https://mlsanity.com/paper/2310.04673/) (2023)
- [FunCodec: A Fundamental, Reproducible and Integrable Open-source Toolkit for Neural Speech Codec](https://mlsanity.com/paper/2309.07405/) (2023)
- [Beyond Turn-Based Interfaces: Synchronous LLMs as Full-Duplex Dialogue Agents](https://mlsanity.com/paper/2409.15594/) (2024)
- [Vocabulary Learning via Optimal Transport for Neural Machine Translation](https://mlsanity.com/paper/2012.15671/) (2020)
- [LARP: Tokenizing Videos with a Learned Autoregressive Generative Prior](https://mlsanity.com/paper/2410.21264/) (2024)
- [SpeechGPT-Gen: Scaling Chain-of-Information Speech Generation](https://mlsanity.com/paper/2401.13527/) (2024)
- [ALMTokenizer: A Low-bitrate and Semantic-rich Audio Codec Tokenizer for Audio Language Modeling](https://mlsanity.com/paper/2504.10344/) (2025)
- [Stack-and-Delay: a new codebook pattern for music generation](https://mlsanity.com/paper/2309.08804/) (2023)
- [Efficient Token Compression for Vision Transformer with Spatial Information Preserved](https://mlsanity.com/paper/2503.23455/) (2025)

## Abstract (arXiv)

Recent advances in speech language models leverage discrete speech representations from pretrained codecs to enable scalable training and generation. However, existing codecs are primarily optimized for compression without accounting for the autoregressive nature of language model training, resulting in suboptimal performance when modeling compressed speech tokens. In this work, we revisit speech discretization from a generative modeling perspective and propose a novel framework that explicitly aligns speech tokenization with autoregressive training. Our approach introduces autoregressive-compatible constraints during codec training, encouraging token sequences that exhibit temporal consistency and predictability. In addition, we propose a heterogeneous downsampling strategy for different layers of speech tokens, distinguishing semantic from acoustic layers, to improve the alignment between semantic tokens and corresponding textual content. Extensive experiments across multiple benchmarks demonstrate that our method bridges the gap between speech compression and generative modeling, enabling more effective continued pretraining of existing language models on speech data. The approach consistently improves performance across multiple codecs, validating its generality and applicability to diverse speech modeling scenarios.