Generate music, create lifelike speech, clone voices, separate stems, repair sections, record, mix, master, and automate full audio workflows. Foundry runs locally on Windows with an NVIDIA GPU, so your scripts, voices, songs, and client audio stay on your machine.

Download for Windows Start free trial Hear examples

No cloud generation / No credit meters / Voice cloning included / Built for private production

50+

Music languages

Speech languages

50 x 5

Emotion control

200+

DSP presets

120+

Commands

Music

Songs, vocals, structure, style, and language control.

Create full tracks, extend ideas, transform references, patch weak sections, and generate multilingual songs locally.

Compare music generators

Speech

Expressive narration that does not sound like filler.

Generate speech in 10 languages, direct emotion line by line, clone voices from short samples, and build multi-speaker scenes.

Explore local AI voice

Studio

A real production workspace, not just a prompt field.

Record, arrange, stem-split, patch, crossfade, process with DSP, mix on a DAW-style timeline, and export final audio.

See what is inside

Local creative agent

Give the boring preparation work to the assistant.

Use the built-in agent for lyrics, music briefs, script preparation, speaker extraction, emotion planning, literature summaries, narration segmentation, batch workflows, and repeatable production tasks.

foundry analyze manuscript.pdf
foundry extract speakers and emotions
foundry generate narration
foundry compose intro music
foundry mix and export

Pick your workflow

YouTubers and faceless channels

Voiceovers, hooks, background music, repeatable production, no per-video credit anxiety.

Game developers

NPC dialogue, character voices, batch generation, ambience, score, and engine-ready export.

Audiobooks and podcasts

Long-form narration, multi-speaker scenes, private manuscripts, music beds, and final export.

Speech studios by language

English Deutsch Français Español Italiano 中文日本語 Русский Português 한국어

Learn fast

Getting started Emotion manager Patching Creative AI agent

Deep dives

Voice cloning ElevenLabs alternative Automation Privacy

Model

This model is hosted for Demodokos Foundry but it can be used for other purposes, enjoy a stable download location and custom quantizations not available elsewhere.

Qwen3-TTS

🤗 Hugging Face | 🤖 ModelScope | 📑 Blog | 📑 Paper | 💻 GitHub

We release Qwen3-TTS, a series of powerful speech generation models developed by Qwen, offering comprehensive support for voice cloning, voice design, ultra-high-quality human-like speech generation, and natural language-based voice control.

Overview

Qwen3-TTS covers 10 major languages (Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian) as well as multiple dialectal voice profiles. Key features:

Powerful Speech Representation: Powered by the self-developed Qwen3-TTS-Tokenizer-12Hz, it achieves efficient acoustic compression and high-dimensional semantic modeling.
Universal End-to-End Architecture: Utilizing a discrete multi-codebook LM architecture to bypass traditional information bottlenecks.
Extreme Low-Latency Streaming Generation: Supports streaming generation with end-to-end synthesis latency as low as 97ms.
Intelligent Voice Control: Supports speech generation driven by natural language instructions for flexible control over timbre, emotion, and prosody.

Quickstart

Environment Setup

Install the qwen-tts Python package from PyPI:

pip install -U qwen-tts

Python Package Usage

python

import torch
import soundfile as sf
from qwen_tts import Qwen3TTSModel

# Load the model
model = Qwen3TTSModel.from_pretrained(
    "Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice",
    device_map="cuda:0",
    dtype=torch.bfloat16,
    attn_implementation="flash_attention_2",
)

# Custom Voice Generation
wavs, sr = model.generate_custom_voice(
    text="其实我真的有发现，我是一个特别善于观察别人情绪的人。",
    language="Chinese",
    speaker="Vivian",
    instruct="用特别愤怒的语气说",
)
sf.write("output.wav", wavs[0], sr)

Evaluation

Zero-shot speech generation on the Seed-TTS test set (Word Error Rate (WER, ↓)):

Model	test-zh	test-en
Qwen3-TTS-12Hz-1.7B-Base	0.77	1.24

Citation

If you find our paper and code useful in your research, please consider giving a star ⭐ and citation 📝:

BibTeX

@article{Qwen3-TTS,
  title={Qwen3-TTS Technical Report},
  author={Hangrui Hu and Xinfa Zhu and Ting He and Dake Guo and Bin Zhang and Xiong Wang and Zhifang Guo and Ziyue Jiang and Hongkun Hao and Zishan Guo and Xinyu Zhang and Pei Zhang and Baosong Yang and Jin Xu and Jingren Zhou and Junyang Lin},
  journal={arXiv preprint arXiv:2601.15621},
  year={2026}
}

cmp-nct/Qwen3-TTS-12Hz-1.7B-VoiceDesign

license: apache-2.0 tag: demodokos

Your GPU is the studio.