Text to Speech
Generate speech from text in a supported language in your browser and export WAV audio.
Quick answer and specifications
The core workflow runs in this browser. Check the supported formats, published limits, and any on-demand model download before you start.
- Inputs
- text
- Outputs
- WAV
- Processing
- On this device, in the browser
- Model languages
- 31 languages — see the language selector
- Last verified
The interface language does not determine the model language. Recognition and pronunciation quality vary; check the result.
Ready to generate speech in this browser.
ONNX Runtime WebGenerated WAV
Audio appears here after generation.
How to convert text to speech locally
Aiviko's Text to Speech (TTS) tool synthesizes natural spoken audio from text in 31 supported languages directly in your browser and exports uncompressed WAV files. Powered by the Supertonic ONNX model running locally via ONNX Runtime Web (WebGPU or WASM fallback), it provides 10 voice styles (5 male, 5 female), adjustable speaking speed, and configurable denoising steps.
1. Choose a supported language and voice style
Select from 31 supported languages (or Auto) and pick one of the 10 on-device voices (Male 1–5 or Female 1–5).
2. Set quality steps, speed, and input text
Choose Fast or High denoising quality and Slow, Normal, or Fast playback speed, then type or paste your text.
3. Generate in browser and download WAV
Run local ONNX synthesis, listen to the result in the built-in player, and download the generated WAV audio.
Supported TTS features and controls
- On-device neural voice synthesis across 31 supported languages including English, Korean, Japanese, Spanish, French, and German
- 10 built-in voice presets (M1–M5 male and F1–F5 female styles) with speed and denoising quality controls
- In-browser ONNX Runtime Web execution supporting WebGPU hardware acceleration and WASM fallback
- Instant in-page audio preview with seek controls and clean WAV file download
Language coverage and synthesis limitations
- The Supertonic model supports 31 languages and does not currently include Chinese (zh) speech synthesis, even when the site UI is viewed in Chinese.
- First-time generation downloads the Supertonic ONNX model files and voice style JSON from the documented model host before caching them in the browser.
- Abbreviations, unusual numbers, loanwords, or mismatched language selections can produce unnatural pronunciation; listen to the full WAV before publishing.
Related speech and audio tools
Frequently asked questions about Text to Speech
Is the text I type sent to an external TTS API?
No. Text normalization, neural synthesis, denoising, and WAV encoding run inside your browser using ONNX Runtime Web. Only the model weights and voice style files are downloaded from the documented host.
Which languages are supported by the voice synthesizer?
The current Supertonic model supports 31 languages including English, Korean, Japanese, Spanish, French, German, Italian, Portuguese, Russian, Arabic, Hindi, and Vietnamese. Chinese synthesis is not supported by this model.
What is the difference between Fast and High quality modes?
Quality controls the number of neural denoising steps during synthesis. Fast finishes more quickly on CPU/WASM, while High runs additional denoising iterations for cleaner audio.
Before you use this tool
Product information reviewed by Aiviko TeamWhere processing happens
The selected input is processed in your browser and is not uploaded to an Aiviko processing server.
What may use the network
The page may download code, fonts, runtime files or machine-learning models. Analytics and advertising load only under the choices and conditions described in the privacy and cookie policies.
Results and limits
Speed, accuracy and maximum practical file size depend on the input, browser and device. Review the guide before using the output for an important decision.