88Z AI, 25 August — Xiaohongshu's FireRed team releases the open-source model FireRedTTS3, based on RedAE and LLM-DiT architectures, supporting zero-shot cloning and precise editing across multiple languages with top benchmark performance.
At a glance
- Supports zero-shot cloning and precise editing across 24 languages and 21 Chinese dialects
- Achieved top results in four public benchmarks, suitable for various speech scenarios
The update
The FireRed team at Xiaohongshu has officially open-sourced FireRedTTS3, a unified model for speech generation and editing. Leveraging RedAE semantic-enhanced continuous representations combined with an LLM-DiT architecture, this single model delivers complex capabilities. It supports zero-shot voice cloning across 24 languages and 21 Chinese dialects, enables natural language-based sound design, and allows for precise editing of specific speech regions. The model achieved top results in four public benchmarks. This advancement offers efficient solutions for scenarios such as audio content production, game dubbing, and brand customer service, demonstrating strong capability and broad applicability in speech processing.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- FireRedTTS3Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.