Use case · Creators & dubbing
Sudanese voice over, without the booth.
Narrate a video, dub an explainer, open a podcast episode or fire off a voice note — in Sudanese Arabic, at two in the morning, with no studio and no voice actor to book. Two AI voices, a commercial licence, and audio you can drop straight into a timeline.
Why this was hard until now
The dialect is the audience
Sudanese creators have had two bad options. Record it yourself — which means a quiet room, a decent mic, retakes, and being on camera-voice every time you publish. Or run the script through a general Arabic TTS and accept a Modern Standard Arabic delivery that lands somewhere between a news bulletin and a school textbook. Neither one sounds like the person your audience follows.
Sudanese Arabic has its own consonants, its own pacing and its own vocabulary, and an audience clocks a foreign delivery in about two seconds. That audience is large and scattered: Sudan itself, plus enormous communities in the UAE, Saudi Arabia, Egypt, the UK and the US who watch Sudanese content precisely because it sounds like home. A voice that sounds like Cairo is not a small compromise — for a lot of viewers it is the reason they scroll past.
What people make with it
YouTube narration
Explainers, history, news commentary, recipe and repair videos — write the script, pick zol or zola, render, cut. Re-take a single line without re-recording the whole thing.
TikTok & Reels
Short-form moves fast and voice over is half the format. A fifteen-second script is a few hundred characters, so a day's worth of posts barely dents a monthly allowance.
Podcast inserts
Intros, outros, sponsor reads, chapter markers and correction notes — the pieces that are tedious to re-record every episode and that listeners expect to sound consistent.
Dubbing & localisation
Bring foreign-language content into Sudanese Arabic line by line, or localise an Egyptian or Gulf original for a Sudanese audience without hiring a full dub cast.
Voice notes
Type it, get a Sudanese voice note, share it on WhatsApp. You can even type in English — useful if you speak the dialect but do not type Arabic at speed.
Poetry & spoken word
The poetry style reads verse line by line with recitation pauses rather than racing through it as prose — a genuinely different delivery, not just a slower one.
The voices
Nobody's voice was taken to make these
Zol (زول) and zola (زولة) are invented voices — blends constructed so they do not reproduce any identifiable speaker. That is a deliberate design choice, not a limitation: it means the audio you publish carries no likeness claim, no unpaid voice actor and no awkward conversation if a video travels further than you expected.
male · warm
Conversational and unhurried. Long-form narration, storytelling, podcast reads.
female · bright
Friendly and clear. Short-form, explainers, announcements and hooks.
The workflow
Script, audition, download
Start on the home page demo with no account at all — paste a line and hear it. Once you sign up, the studio playground is where the actual work happens: audition a paragraph, switch voices, try a style, and download the takes you want as WAV or MP3. If you are producing at volume, the same synthesis is one POST request away in the API, so a script file can become a folder of numbered audio clips in a single loop.
Two practical habits: render long content in short chunks so you get clean edit points and cheap re-takes, and remember that the first request after a quiet spell can take up to a minute and a half while a GPU wakes — after that it is seconds. Free accounts get 5,000 characters a month and 300 per request, which is plenty for auditioning; Starter raises that to 100,000 a month with the commercial licence attached.
Building something that answers the phone rather than plays in a feed? The IVR and call center page covers telephony formats and latency, and the Sudanese text-to-speech overview explains how the model was built.
Questions
Can I use Sautna audio in monetised YouTube or TikTok videos?
Yes, on any paid plan. Starter at $9/month includes a commercial licence covering the audio you generate, including monetised video, client work and ads. The free tier is for personal use and trying things out.
Whose voice is it? Do I need anyone's permission?
Nobody's. Zol and zola are invented voices — blends built so they do not reproduce any identifiable person. There is no voice actor to credit, no estate to clear and no likeness dispute waiting for you when a video does well.
Can I dub an existing video into Sudanese Arabic?
Yes. Translate or rewrite your script into Sudanese Arabic, render it line by line, and cut the audio against your timeline. Rendering in short chunks rather than one long block gives you clean edit points and lets you re-take a single line without regenerating everything.
Does it handle Sudanese poetry and songs lyrics?
It reads verse the way it reads prose — clearly, in dialect. Put each line on its own line and add punctuation where you want the pauses. It is a speech model, not a singing model, so melody is out of scope.
How much audio do 100,000 characters get me?
Roughly two to three hours of speech, depending on pacing — Arabic averages a little over ten characters per second of audio. That is the Starter plan's monthly allowance, which covers a weekly podcast episode or a steady run of short-form videos.
Can I type in English and get Sudanese audio?
For voice notes, yes — the note tool takes English input and speaks it back in Sudanese Arabic, which is handy for diaspora users who speak the dialect but do not type Arabic quickly. For production work, writing the script in Arabic gives you far more control over the result.
Voice your next video in Sudanese
5,000 characters a month, free, no card. Add the commercial licence for $9/month whenever the work starts paying.
Start free