DS Speech
Voice Cloning
Upload or record your audioClick to upload or drag a file here. MP3 / WAV / M4A, up to 10MB, 10–60 seconds long.
.MP3.WAV.M4A<10MB10~60s

Online voice cloning

Create a custom AI voice with Qwen-Audio-3.0-TTS

Voice cloning captures characteristics from authorized reference audio and creates a reusable voice for text-to-speech tasks. DS Speech combines upload, review, voice management, and speech generation in one online workspace.

How to clone a voice online

Clear, continuous single-speaker audio improves the reliability of voice review and later speech generation.

  1. 1

    Upload or record reference audio

    Provide 10–60 seconds of clear speech without music, background noise, other speakers, or long pauses.

  2. 2

    Add voice details and submit

    Set a name, description, and avatar. Track the review status from My Voices after submission.

  3. 3

    Use the approved voice

    Once approved, select the voice in Text to Speech for new scripts, languages, and delivery styles.

Content that benefits from voice cloning

A reusable voice keeps a series consistent, but you must obtain legal permission before cloning another person's voice.

Video series and podcasts

Reuse one voice after script changes and reduce repeat recording sessions.

Courses and spoken content

Maintain a consistent narrator across lessons, chapters, and long-running shows.

Characters and brand voices

Build reusable voices for game characters, product narration, and branded content.