Qwen Voice Design
Qwen Voice Design — Write a persona brief, pick from 10 languages
Write a persona brief in English or Chinese, then pick the spoken language from a menu of 10. Hear an instant preview, then save the voice ID for Qwen TTS scripts and reference-audio video workflows. Credit pricing, cost shown before you generate.
Persona briefs
Written in English or Chinese
Ten languages
Pick one at design time
Instant preview
Sample clip on every design
Reusable voice ID
Call it from Qwen TTS later
Video reference audio
Custom speech for video workflows
Model Overview
Qwen Voice Design
Qwen Voice Design builds a speaking voice from persona notes alone. Write a free-form brief of up to 2,048 characters in English or Chinese — gender, age, pitch, pace, emotion, use case — then pick the language the finished voice should speak: the menu covers 10 languages, from English and Chinese to Japanese, Korean, French, and Russian.
The details
Each run returns a preview clip of your test lines plus a voice identifier. Pass that ID to Qwen TTS and every later script — narration, dialogue, promo reads — is spoken by the saved voice without re-designing it, and the preview audio can feed reference-audio video workflows.
Designs run on Moosky credits with the cost shown before you generate. If you already hold a recording of a real person, skip the brief: CosyVoice Voice Clone learns that timbre from a 5-20 second sample.
Why use it on Moosky AI?
Model-specific controls with no subscription, clean outputs, and a workflow built for fast creative iteration.
Persona briefs, not sliders
Describe gender, age, pitch, pace, emotion, and use case in your own words — up to 2,048 characters.
A 10-language menu
Pick one of 10 languages at design time: English, Chinese, German, Italian, Portuguese, Spanish, Japanese, Korean, French, or Russian.
Sample before you save
The preview speaks your test lines right away, so you can reject weak designs before committing to an ID.
Reusable voice ID for Qwen TTS
The saved ID drives every later Qwen TTS synthesis and any reference-audio video workflow that takes custom speech.
How it works
Three focused steps from setup to finished output.
Credit-based
Simple, honest pricing
Uses Moosky credits with cost shown before you generate. No subscription required.
More from Moosky AI
Explore our full suite of AI tools.
Frequently Asked Questions
Which languages can I pick for a designed Qwen voice?
The language menu covers 10: English, Chinese, German, Italian, Portuguese, Spanish, Japanese, Korean, French, and Russian. You choose it at design time; the persona brief itself is written in English or Chinese.
What goes into a good voice description?
Name several dimensions at once — gender, age, pitch, pace, emotion, and the job the voice will do. Vague adjectives and celebrity imitation produce mush; specific persona notes give the model something to speak.
Do I need to record any audio?
No. Voice Design invents a voice from written persona notes alone. If you have a clip of a real person instead, CosyVoice Voice Clone learns that timbre from a 5-20 second sample.
What do I keep after designing a voice?
A preview audio clip and a voice identifier. Call that ID from Qwen TTS for any later script, or feed the preview into video workflows that accept reference audio.
How is Qwen voice design priced?
Each design runs on Moosky credits with the cost shown before you generate — no subscription. Synthesis in the saved voice bills separately through Qwen TTS.