Qwen3-TTS CustomVoice API

Alibaba Cloud, open modelqwen3-tts-customvoiceModel source: Hugging Face

Reads your text aloud in the voice you choose.

Qwen3-TTS CustomVoice turns text into natural speech with one of its nine ready-made voices, and lets you direct the tone and emotion with plain instructions. It costs €19.00 per million characters, with the same /v1/audio/speech call as the OpenAI API.

What it does

Text to speech

tts

How it is charged

Each million characters

19.00 €

Where it runs

Spain (EU)

MIG B200

Brain size

Small and fast

TTS, Preset voices

What it is good at

A consistent voice for your brand. Pick one of the nine voices and use it everywhere: assistant, IVR, videos, audiobooks. Consistent across content and time.
Tone and emotion on demand. Describe how it should sound in instructions (up to 500 characters): calm, cheerful, serious.
Natural prosody. Pauses, emphasis and intonation that sound spoken, not read.
Spanish and nine more languages. Chinese, English, Japanese, Korean, German, French, Russian, Portuguese, Spanish and Italian. Every voice speaks all ten.

It does not clone voices. To speak with a voice from your own sample, use Qwen3-TTS Base.

Pricing

€ per million characters, VAT included

Each million characters (use)19.00 €

try it with 5 € freeActivated with your card, and you only pay for what you use.

For your technical team

You send the text and get the audio back. You pick the model and the voice.

curl https://api.gpuflow.ai/v1/audio/speech \
  -H "Authorization: Bearer $GPUFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "model": "qwen3-tts-customvoice",
        "input": "Hola",
        "voice": "Ryan" }' --output speech.wav

Limits and formats

Requests per minute: 1,000 per key, with bursts up to double; raisable to 5,000 from the dashboard.
Output formats: wav (default), mp3, flac, pcm, aac and opus.
Voices: Nine ready-made voices that speak ten languages; tone is set with instructions.

Qwen3-TTS CustomVoice compared with similar models

	Qwen3-TTS CustomVoice	Qwen3-TTS Base (clone)
Best for	Ready-made voices with adjustable tone and emotion	Your own voice from a short sample
Voices	Nine ready-made voices	Clones your sample
Price	19.00 € per million characters, VAT included	5.00 € per million characters, VAT included

A ready-made voice with control over tone: Qwen3-TTS CustomVoice; your own voice from a sample: Qwen3-TTS Base. Both use the same endpoint.

Frequently asked questions

How much does Qwen3-TTS CustomVoice cost?

€19.00 per million characters, VAT included. No monthly fee.

Where does it run?

On our NVIDIA B200 GPUs, in a data centre in Spain.

Is the text-to-speech API compatible with OpenAI's?

Yes, same /v1/audio/speech call with model, input and voice. It returns wav by default, or mp3, flac, pcm, aac and opus.

Which voices are there?

Nine: Vivian (the default), Serena, Uncle_Fu, Dylan, Eric, Ryan, Aiden, Ono_Anna and Sohee. Each has a native language, but any of them speaks all ten supported languages.

Can I clone my own voice?

Not with this model. Qwen3-TTS Base clones a voice from a 1 to 30 second sample and its transcript, on the same endpoint.

Is my text stored?

No. The text is processed in memory and the audio comes back in the same call; we do not keep either, and they are not used for training.