Hamsa
Arabic voice API sorted by Gulf dialect codes, deployable inside Saudi
Visit website →About
Hamsa is an Arabic-first voice platform for developers, with text-to-speech, transcription and voice cloning behind one API. It is the only vendor in our guide whose documentation sorts voices by dialect code rather than plain "Arabic".
The voice list carries a Saudi code with voices such as Hiba and Fahd, a separate Gulf code with Khalid and Rahma, and distinct codes for Emirati, Qatari, Kuwaiti, Omani and Bahraini voices, next to Levantine, Egyptian, Iraqi and Modern Standard entries. The documentation states that diacritics are honoured and that mixed Arabic-English text is supported, two claims most competitors do not make in writing. In July 2026 Hamsa announced a sovereign deployment inside Saudi Arabia through the local provider OmniOps, so Kingdom-resident processing is available on request. We found no published price list, no stated headquarters and no public page on cloning consent; the arena comparison that includes Hamsa has no results yet.
Pricing in November 2026: not published; contact the company.
AI-assisted draft, human-reviewed before publishing — see how we choose & review tools.
Why we picked it
Hamsa is the one vendor that distinguishes a Saudi voice from a generic Gulf one in its own voice codes and documents diacritics and code-switching, and it can be deployed inside Saudi Arabia. For a guide about which voices are actually labelled Saudi, that combination had to be in the catalog.
Featured in our guides
Related tools
- 1 Orpheus Arabic Saudi (on Groq) Six Saudi-dialect voices on Groq at $40 per million characters, in preview Paid
- 2 SILMA AI Arabic-first speech: open base model, auto-diacritics and a Saudi option Paid
- 3 Nabrah Riyadh text-to-speech with ten Saudi and Gulf voices, priced in riyals Freemium
- 4 Speechify Read-aloud app for documents and web pages with a handful of Arabic voices Freemium
- 5 Murf AI Studio voice-over tool with forty-plus Arabic voices and no dialect label Paid
- 6 OpenAI Text-to-Speech English-first voices that read Arabic; current models retire January 2027 Paid