Step-Audio-EditX Technical Report

TL;DR

Step-Audio-EditX is an open-source LLM-based audio model that enables expressive, iterative audio editing and zero-shot TTS using synthetic data, surpassing previous models in emotion and control tasks.

Contribution

It introduces a novel large-margin learning approach that eliminates the need for embedding priors, enhancing voice expressivity and control in audio editing.

Findings

Outperforms MiniMax-2.6-hd and Doubao-Seed-TTS-2.0 in emotion editing.

Enables high expressivity and iterative control without auxiliary modules.

Demonstrates robust zero-shot text-to-speech capabilities.

Abstract

We present Step-Audio-EditX, the first open-source LLM-based audio model excelling at expressive and iterative audio editing encompassing emotion, speaking style, and paralinguistics alongside robust zero-shot text-to-speech (TTS) capabilities. Our core innovation lies in leveraging only large-margin synthetic data, which circumvents the need for embedding-based priors or auxiliary modules. This large-margin learning approach enables both iterative control and high expressivity across voices, and represents a fundamental pivot from the conventional focus on representation-level disentanglement. Evaluation results demonstrate that Step-Audio-EditX surpasses both MiniMax-2.6-hd and Doubao-Seed-TTS-2.0 in emotion editing and other fine-grained control tasks.