SONORA · VOICE STUDIO
How it works & FAQ
Learn how Sonora Voice Studio turns scripts into expressive AI voice narrations, how per-request word limits work, and how to estimate generation time and export audio.
The studio is interactive, so JavaScript is required to generate, preview, and export audio. This guide remains readable without it.
HOW IT WORKS
From script to narration in three steps
Sonora turns your written script into an AI voice narration while keeping the generation and export controls in the studio.
- Write and shape. Paste or write a script, then add optional emotion and delivery tags to guide the reading.
- Choose and preview. Select an engine and voice, check the approximate generation time, generate the narration, and listen to the completed audio.
- Export. Guests can download MP3. Sign in with Google to enable WAV, FLAC, AAC, and OGG as well.
COMMON QUESTIONS
Frequently asked questions
How do the word limits work?
Guests can submit up to 50 words per request with Low Quality. After creating a local account, each request can contain up to 1,000 words with Low Quality, 500 with Balanced, or 250 with Enhanced Thinking. Verified admins are exempt from these word limits; character and audio safety limits still apply. Whitespace-separated words in the script and delivery tags both count. There is no daily word or request quota. Short-window anti-abuse and server-capacity safeguards still apply.
Which models can I use?
Guests can use Low Quality. Creating a local account enables Low Quality, Balanced, and Enhanced Thinking. No subscription is required.
Which audio formats can I export?
Guests can download MP3. Creating a local account enables all five formats: MP3, WAV, FLAC, AAC, and OGG.
How long will generation take?
An approximate generation time appears above Generate audio before you start. It depends on input length and the selected model, not measured provider performance. A live elapsed timer runs during generation. Actual time varies with language, delivery, network, and provider load.
Will tags produce an exact voice delivery?
Tags guide delivery, but generated output is not guaranteed to match the requested delivery exactly.
Can I generate, preview, or export audio without JavaScript?
JavaScript is required for generation, preview, and export.