Skip to content
SPECIFICTION
Get listed

Qwen3-TTS

github.com

Free plan WebLinuxAPISelf-hosted
Visit github.com
Spec score 8.9 #20 of 233 in Text-to-speech tools

Spec sheet

What it does
Qwen3-TTS is an open-source series of text-to-speech models for speech generation, voice design, and voice cloning.Read 7 Oct 2026 · github.com
Product
Qwen3-TTS is an open-source series of text-to-speech models developed by the Qwen team at Alibaba Cloud.Read 7 Oct 2026 · github.com
Platforms
web,api,self_hostedRead 20 Sep 2026 · github.com
License
The GitHub repository lists the Apache-2.0 license.Read 7 Oct 2026 · github.com
API access
YesRead 20 Sep 2026 · github.com
API integration
The project links Alibaba Cloud DashScope real-time APIs for custom voice, voice cloning, and voice design.Read 7 Oct 2026 · github.com
API pricing
Alibaba Cloud lists international Qwen3-TTS API rates of $0.115 per 10,000 input characters for Voice Design and Voice Cloning, with output free; the page also lists a 110,000-character free quota for each.Read 7 Oct 2026 · alibabacloud.com
API requirements
The linked Alibaba Cloud real-time synthesis guide says API use requires configuring an API key and installing the latest DashScope SDK when using that SDK.Read 7 Oct 2026 · help.aliyun.com
Commercial use
YesRead 20 Sep 2026 · github.com
Download and install
Models can be downloaded through Hugging Face or ModelScope, and the project provides a `qwen-tts` Python package installable from PyPI.Read 7 Oct 2026 · github.com
Export formats
WAVRead 20 Sep 2026 · github.com
Generation modes
The models support streaming and non-streaming speech generation, voice design, voice cloning, and natural-language voice control.Read 7 Oct 2026 · github.com
Hardware and setup
The README demonstrates model loading on CUDA and recommends FlashAttention 2 to reduce GPU memory use, subject to compatible hardware.Read 7 Oct 2026 · github.com
Inference integration
The README states that vLLM-Omni supports offline Qwen3-TTS inference and that online serving is planned for later support.Read 7 Oct 2026 · github.com
Languages
10Read 20 Sep 2026 · github.com
Languages
The project lists support for Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian, as well as multiple dialectal voice profiles.Read 7 Oct 2026 · github.com
Latency
The README says the streaming architecture can produce the first audio packet after one character and reports end-to-end synthesis latency as low as 97 ms.Read 7 Oct 2026 · github.com
Local installation
The project provides a Python package install and says it recommends a fresh Python 3.12 environment.Read 7 Oct 2026 · github.com
Local web interface
A local Gradio web UI demo can be launched with the qwen-tts-demo command.Read 7 Oct 2026 · github.com
Model hosting
The README links model downloads on Hugging Face and ModelScope and documents loading weights locally or by model ID.Read 7 Oct 2026 · github.com
Model sizes
Released models include 0.6B and 1.7B parameter variants for custom voice or base tasks, plus a 1.7B voice design model.Read 7 Oct 2026 · github.com
Serving limitation
The README says vLLM-Omni supports offline inference currently, with online serving planned for later.Read 7 Oct 2026 · github.com
Streaming
The README says the models support streaming and non-streaming generation and reports end-to-end synthesis latency as low as 97 ms.Read 7 Oct 2026 · github.com
Voice cloning
The Base models can clone a voice from a reference audio clip and are described as capable of rapid cloning from three seconds of input audio.Read 7 Oct 2026 · github.com
Voice cloning input
Voice cloning uses reference audio and its transcript; the README says using only the speaker embedding without a transcript may reduce cloning quality.Read 7 Oct 2026 · github.com
Voice options
The released models support custom voices, voice design from natural-language descriptions, and voice cloning from reference audio.Read 7 Oct 2026 · github.com
Web demo
The project includes a local web UI demo launched with `qwen-tts-demo`.Read 7 Oct 2026 · github.com

Plans and prices

Qwen3-TTS open-source models Free The project is released under the Apache-2.0 license; no price is stated. 0.6B and 1.7B models · 10 languages Read 7 Oct 2026 · github.com

Every Qwen3-TTS plan and price →

Where Specifiction places it

Alternatives to Qwen3-TTSAll alternatives →