Alibaba Cloud: Qwen3-TTS-Flash

Qwen3 TTS Flash converts prepared text into spoken audio with a selected voice. It suits multilingual narration and spoken responses where the script is known before synthesis begins.

Compare models

qwen3-tts-flash

AvailableAlibaba CloudText to speechSync

Input / Output: $10.00 / $0.00
Modalities: Audio

About Qwen3-TTS-Flash

Qwen3-TTS-Flash is Alibaba's text-to-speech model with ready-made system voices. You send prepared text and a voice name over an HTTP call and receive spoken audio in WAV format. It speaks Mandarin, English, and a range of other languages, plus several Chinese dialects. It does not clone voices or follow style instructions like CosyVoice v3.5 Plus, and it waits for the complete text, unlike the Realtime variant.

Where it works well

Offers system voices, so you can start without recording or designing a voice.
Speaks Chinese dialects such as Beijing, Shanghai, Sichuan, Nanjing, Shaanxi, Hokkien, Tianjin, and Cantonese.
Also covers English, German, Italian, Portuguese, Spanish, Japanese, Korean, French, and Russian.
Uses a plain HTTP API and returns WAV output, which is simple to integrate.

When to choose another model

It cannot clone a voice or design a new one; choose CosyVoice v3.5 Plus for custom voices.
It offers no instruction control for emotion or style beyond the voice you pick.
It needs the full text before synthesis; use the Realtime variant to start speaking sooner.

Getting started

Create API key
Create a key in Console, then use it with every model on the platform.

Send your first request

Copy the example for your language and run it against the endpoint.

Text to speechTokenLab endpoint

POST/v1/audio/speech

curl -X POST "https://api.tokenlab.sh/v1/audio/speech" \
  -H "Authorization: Bearer sk-xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "operation": "tts",
  "response_format": "wav",
  "speed": 1,
  "voice": "Cherry",
  "model": "qwen3-tts-flash",
  "input": "Hello from TokenLab."
}'

Parameters

Text to speech

Parameter	Values and limits	Default	Notes
`input`	Not provided	Not provided	Not provided
`language_code`	Not provided	Not provided	This field is accepted. The current public contract does not specify a complete value range or list. No additional values are promised. The omitted-field default is not verified.
`response_format`	wav	Not provided	Output is WAV. This accepted compatibility field does not transcode output. Use the response Content-Type and file extension. The omitted-field default is not verified.
`speed`	Not provided	Not provided	The gateway accepts numbers from 0.25 to 4.0. The selected supplier can impose a narrower range. A uniform omitted-field default is not verified.
`voice`	Cherry, Serena, Ethan, Chelsie, Momo, Vivian, Moon, Maia, Kai, Nofish, Bella, Jennifer, Ryan, Katerina, Aiden, Eldric Sage, Mia, Mochi, Bellona, Vincent, Bunny, Neil, Elias, Arthur, Nini, Seren, Pip, Stella, Bodega, Sonrisa, Alek, Dolce, Sohee, Ono Anna, Lenn, Emilien, Andre, Radio Gol, Jada, Dylan, Li, Marcus, Roy, Peter, Sunny, Eric, Rocky, Kiki	`Cherry`	Not provided

Speech output is WAV audio. The response_format parameter remains accepted for request compatibility and does not transcode this output; use the response Content-Type and file extension for the actual audio format. The default voice is Cherry.

Pricing

The Verified price applies to Verified, which costs less on most models. The Official price is the model maker's published price and applies to the more reliable Official route. Auto bills the route that completes the request.

Cosy Tts Number

per 1M characters

Official price: Input $10.00 / Output $0.00
Verified price: —
Discount: —

	Official priceper 1M characters	Verified priceper 1M characters	Discount
Cosy Tts Number	Input $10.00 / Output $0.00	—	—

Usage & activity

Success rate is the share of requests that completed. Latency is how long a full response takes; P95 means 95% of requests finished within that time.

Usage & availability

Last 24 hours

Requests
Success rate
P95 latency
Total tokens

Model performance

Last updated Nov 1, 2026, 8:13 AM UTC

Metrics appear once privacy and data volume thresholds are met.

Data is based on aggregate user requests, excluding status checks.

Open in Console

Open Qwen3-TTS-Flash in Console with a prompt ready to edit or send.

Help me try qwen3-tts-flash with a short audio request at /v1/audio/speech. Show the result and cost.

Use cases

Narrated articles
Turn written pages into audio with a chosen system voice for a listen-along feature.
Spoken app responses
Read assistant answers aloud in the user's language once the text is final.
Multilingual voiceovers
Generate the same script in several languages for short videos and explainers.

Prompt examples

Read this product description in Spanish with a clear, neutral voice.

Convert the following news summary to speech in Mandarin.

Say 'Your order has shipped' in Japanese.

FAQ

What exactly is Qwen3-TTS-Flash?

A text-to-speech model from Alibaba that converts prepared text into audio using a system voice. Requests go over HTTP and the audio comes back in WAV format.

Which languages does it speak?

Chinese (Mandarin and several dialects), English, German, Italian, Portuguese, Spanish, Japanese, Korean, French, and Russian. The voice list shows which voices suit which language, so check it first.

Can I clone a voice with it?

No. It supports system voices only, with no cloning, voice design, or instruction control over style. For those features, use CosyVoice v3.5 Plus, which works with custom voices.

How is it different from Qwen3-TTS-Flash-Realtime?

This model takes complete text in one request. The Realtime model streams over WebSocket and can start speaking while text is still arriving, which suits conversational assistants.

What audio format does it return?

WAV, which plays directly in most players and audio editors. Because the whole text is synthesized in one request, it suits finished narration, voiceovers and app responses rather than live conversation.

How much does Qwen3-TTS-Flash cost?

On TokenLab, Qwen3-TTS-Flash costs Input $10.00 / Output $0.00 per 1M characters. The pricing table above shows the full breakdown. Rates depend on the billing unit, specification, and usage. Compare matching conditions in the model's detailed pricing; a single rate does not determine the total cost.

Which endpoint should Qwen3-TTS-Flash use?

Use https://api.tokenlab.sh/v1/audio/speech for Qwen3-TTS-Flash. The request example below shows the matching code shape.

Which operations does Qwen3-TTS-Flash support?

Qwen3-TTS-Flash supports Text to speech. Select an operation above to see its endpoint and request example.