Fish-Speech: fast/slow autoregression with grouped finite-scalar audio codes

What changed

Integrated a serial fast/slow autoregressive design and grouped finite-scalar audio quantization with language-model linguistic processing for multilingual speech synthesis.

Scientific scope

Integrated a serial fast/slow autoregressive design and grouped finite-scalar audio quantization with language-model linguistic processing for multilingual speech synthesis. Inherits LLMs, hierarchical token prediction, finite-scalar quantization and neural vocoders. No general guarantee of100percent codebook use, arbitrary polyphone correctness or universal voice cloning; original method is distinct from later closed product claims.

Evidence

The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits LLMs, hierarchical token prediction, finite-scalar quantization and neural vocoders. No general guarantee of100percent codebook use, arbitrary polyphone correctness or universal voice cloning; original method is distinct from later closed product claims.

Contributors

• Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.

• Fish-Speech: Leveraging Large Language Models for Advanced Multilingual Text-to-Speech Synthesis

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction