Text to Speech (Faseeh)

Wittify’s Arabic-first voice synthesis engine. Turn text into natural speech, clone your own voice, or design a new one from scratch.

Faseeh turns text into natural speech. Pick a voice, type or paste your script, tune speed and clarity, and download the audio. Clone your own voice from a short sample, or design a brand-new one by describing the attributes you want.

What you get

Arabic-first voices

Hundreds of Arabic voices across 21 dialects. Plus English (10 accents), Chinese (15 varieties), and 48 more language clusters with 2-17 dialects each.

Studio workspace

A two-column workspace: voice and language picker on the start edge, text editor and Generation Settings on the other side. Live history of every project on the sidebar.

Voice cloning

Upload or record a 3-10 second sample, transcribe it, and clone your own voice. Cloned voices are private to your account.

Voice design

Design a new voice by picking attributes (age, gender, accent, energy). A two-step wizard, no audio sample needed.

Start here

Studio

The main workspace. Pick a voice, write your text, click Generate to make audio.

Voice Picker

The full voice catalog with Explore, My Voices, Favorites, and All Voices tabs.

Clone Wizard

Make a voice that sounds like you, in two steps.

Design Wizard

Build a brand-new voice from attribute dropdowns.

Pages in this system

Studio

The main workspace. Pick a voice, write or paste up to ~5,000 characters, set Speed, Quality, Similarity, Denoise, Preprocess Prompt, Postprocess Output, and click Generate.

Voice Library

Your custom (cloned + designed) voices. Preview, rename, or delete. Each voice is private to your account.

Voice Picker

Full catalog with Explore, My Voices, Favorites, and All Voices tabs, a language → dialect picker, attribute filters, and a Create A Voice entry that opens the cloning or design wizard.

Clone Wizard

Two steps: upload or record a sample (3-10 seconds), transcribe + name your voice, save to your library.

Design Wizard

Two steps: pick attribute dropdowns (language, gender, age, accent), generate a sample, save to your library.

History

Every TTS project you generate is saved automatically. Click any row in the sidebar to reload the editor with the original text and voice.

Generation Settings (the six controls)

The right-side Generation Settings panel in the Studio exposes six controls. These are the only controls in the panel.

Setting	What it does
Speed	Adjust the pace of speech. Slower for narration, faster for ads.
Quality	Move toward speed for faster generation, or toward quality for higher fidelity.
Similarity	Lower for stable, consistent delivery. Higher for closer match to the reference voice.
Denoise	Removes background noise from cloned voices.
Preprocess Prompt	Trims silence on the reference audio and adds end-of-sentence punctuation to the reference text if missing.
Postprocess Output	Removes long silences from the generated audio.

Common questions

Are Faseeh voices tied to a specific external engine?

No. Every Faseeh voice is a Wittify voice. The engine that powers each voice is an internal routing detail and is not surfaced anywhere in the UI.

Can my teammates use my cloned voices?

No. Cloned and designed voices are scoped to your account. To share a voice with another user, that user has to clone or design their own copy.

How long does generation take?

For short text and the default settings, less than 1 second. Longer text or pulling Quality toward higher fidelity takes longer. The button shows a spinner while it’s working, you’ll get the audio link as soon as it’s done.

Where do my generated projects go?

They’re saved automatically to your History sidebar. Click any row to load it back into the editor with the original text and voice settings, then make tweaks and regenerate.

Can I clone a celebrity's voice?

Only voices you have permission to clone (your own, or one a person has explicitly authorised). Wittify enforces a consent step in the Clone Wizard.

My text is in mixed Arabic and English. Will it work?

Yes. The voice will switch between languages naturally if the underlying voice supports both. Pick a voice tagged for both languages in the Voice Picker.