Beta
Arabic speech, in your dialect
Note: The model is still being trained, so output quality may vary. For best results, use longer utterances.
Regional Arabic voices
Gulf, Levantine, Egyptian, Iraqi, Sudanese, Yemeni, Maghrebi, and MSA.
Clone a voice
Upload a short voice sample to make Sawtak Arabi speak in that same voice and dialect.
Streaming audio
Audio arrives while the rest is still being generated.
OpenAI-compatible
Point an existing SDK at our base URL. Nothing else changes.
Connect to your voice AI stack.
Use Arabic speech in the frameworks and real-time pipelines your team already builds with, through the OpenAI-compatible endpoints you already call.
Audiobooks & narration
Narrate books, articles and documentaries in the dialect the story is set in, one voice per character.
E-learning
Voice every lesson, drill and pronunciation exercise with clear, consistent audio from one key.
Media & dubbing
Prototype dubbing and localisation pipelines with per-character voices and streaming playback.
Pay for what you generate.
New accounts get $5 of credit once their email is verified.
Text to Speech
$50
per 1M charactersBilled per character, spaces and diacritics included.
Speech to Text
$0.50
per hour of audioPublished per hour, billed per second of audio.
Run it in your own cloud, or on-premises.
Both engines deployed on your infrastructure. You decide where audio lives, who can reach it, and how it scales.
FAQs
The things developers ask before they write the first request.
Something not covered here? Get in touch
Let’s talk about your project.
Book a 30-minute call with Mohammed to discuss how Sawtak Arabi can help.