Qwen3-TTS
Qwen3-TTS is an open-source text-to-speech model series from Alibaba Cloud’s Qwen team.
What Qwen3-TTS is
Qwen3-TTS is an open-source text-to-speech model series from Alibaba Cloud’s Qwen team. It generates speech in streaming or non-streaming mode and supports voice design from natural-language descriptions, voice cloning from reference audio, and natural-language voice control. The released models include 0.6B and 1.7B parameter variants for custom voice or base tasks, plus a 1.7B voice design model. The listed language count is 10, and WAV is a supported export format. The project provides a Python package and model weights that can be loaded locally from Hugging Face or ModelScope, along with a local Gradio interface launched using `qwen-tts-demo`. It recommends a fresh Python 3.12 environment and demonstrates CUDA model loading; FlashAttention 2 is recommended to reduce GPU memory use on compatible hardware. The README reports synthesis latency as low as 97 ms. Alibaba Cloud DashScope offers linked real-time APIs for custom voice, cloning, and voice design; API use requires an API key. The project uses the Apache-2.0 license and permits commercial use. Its vLLM-Omni integration currently supports offline inference; online serving is planned for later.
Good
- Open-source Apache-2.0 license permits commercial use.
- Supports streaming and non-streaming speech generation.
- Includes voice design and voice cloning.
- Provides a local Gradio web demo.
- Models can be loaded locally from linked repositories.
Watch out
- Voice cloning uses reference audio and its transcript.
- Local CUDA loading is demonstrated; compatible hardware matters.
- vLLM-Omni online serving is planned, not currently supported.
- Hosted API use requires an API key.
Who it suits
Qwen3-TTS suits developers or teams seeking open-source speech generation models with voice design or cloning. Local use requires working with the project's Python installation and model-loading setup.
Verdict
Qwen3-TTS offers several speech-generation modes and can be used locally or through linked DashScope APIs. The documented vLLM-Omni support is offline only, so online serving through that integration is not yet available.
Plans and prices
From Qwen3-TTS's own pricing page, checked 7 Oct 2026.
| Plan | Price | What it gives you |
|---|---|---|
| Qwen3-TTS open-source models | Free | 0.6B and 1.7B models · 10 languages · The project is released under the Apache-2.0 license; no price is stated. |
The facts
Each one read from Qwen3-TTS's own pages, with the date it was checked.
What it does
- API access
- Yeschecked 20 Sep 2026
- API integration
- The project links Alibaba Cloud DashScope real-time APIs for custom voice, voice cloning, and voice design.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- Commercial use
- Yeschecked 20 Sep 2026
- Download and install
- Models can be downloaded through Hugging Face or ModelScope, and the project provides a `qwen-tts` Python package installable from PyPI.checked 7 Oct 2026
- Export formats
- WAVchecked 20 Sep 2026
- Generation modes
- The models support streaming and non-streaming speech generation, voice design, voice cloning, and natural-language voice control.checked 7 Oct 2026
- Hardware and setup
- The README demonstrates model loading on CUDA and recommends FlashAttention 2 to reduce GPU memory use, subject to compatible hardware.checked 7 Oct 2026
- Inference integration
- The README states that vLLM-Omni supports offline Qwen3-TTS inference and that online serving is planned for later support.checked 7 Oct 2026
- Languages
- The project lists support for Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish, and Italian, as well as multiple dialectal voice profiles.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- License
- The GitHub repository lists the Apache-2.0 license.checked 7 Oct 2026
- Local installation
- The project provides a Python package install and says it recommends a fresh Python 3.12 environment.checked 7 Oct 2026
- Local web interface
- A local Gradio web UI demo can be launched with the qwen-tts-demo command.checked 7 Oct 2026
- Model hosting
- The README links model downloads on Hugging Face and ModelScope and documents loading weights locally or by model ID.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- Product
- Qwen3-TTS is an open-source series of text-to-speech models developed by the Qwen team at Alibaba Cloud.checked 7 Oct 2026
- Serving limitation
- The README says vLLM-Omni supports offline inference currently, with online serving planned for later.checked 7 Oct 2026
- Streaming
- The README says the models support streaming and non-streaming generation and reports end-to-end synthesis latency as low as 97 ms.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- 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.checked 7 Oct 2026
- Voice options
- The released models support custom voices, voice design from natural-language descriptions, and voice cloning from reference audio.checked 7 Oct 2026
Where it runs
- Platforms
- web,api,self_hostedchecked 20 Sep 2026
Alternatives to Qwen3-TTS
- 1
Voicebox
Free planOpen sourceAndroidiOSWebAPI9.3From $12 a year - 2
Speechify
Free planAndroidiOSWebAPI9.0From $10 a month - 3
Resemble AI
Free planAndroidiOSWebAPI9.0From $280 a month - 4
iSpeech
Free planOpen sourceAndroidiOSWebAPI9.0Free plan - 5
Fliki
Free planAndroidiOSWebAPI8.8From $28 a month - 6
Voice.ai
Free planAndroidiOSWebAPI8.8From $5 a month