Arabic TTS Models

SILMA Arabic TTS Models are state-of-the-art Arabic Text-to-Speech models offering natural-sounding synthesis for Modern Standard Arabic (Fus'ha) and diverse regional dialects

Try our TTS Now

Compare to Other Models

SILMA TTS v2.0

<300 ms TTFT

Low streaming latency with sub-300 ms TTFT - excluding network

Rock-solid Performance

Superior stability, audio quality and naturalness

On-prem deployment

Flexible hosting options for sensitive customers

2-6x Lower Cost

Scale your voice agent platform cost-effectively. Cut yourAPI costs by 2-6x

Bilingual Fluidity

Native, code-switching support for Saudi Najdi, Modern Standard Arabic (Fusha), and English within a single, unified pipeline

Tailored for Arabic

Where others generalize, our models specialize, offering unmatched depth in dialectal pronunciation

Control

Ability to control style variance and speed

Voice cloning

Our models support high quality voice cloning

Voices

8 brand-new voices

2-6x Lower Cost

Scale your voice agent platform cost-effectively. Cut yourAPI costs by 2-6x

Bilingual Fluidity

Native, code-switching support for Saudi Najdi, Modern Standard Arabic (Fusha), and English within a single, unified pipeline

Tailored for Arabic

Where others generalize, our models specialize, offering unmatched depth in dialectal pronunciation

Control

Ability to control style variance and speed

Voice cloning

Our models support high quality voice cloning

Voices

8 brand-new voices

Open-source Models

Modern Standard Arabic
SILMA TTS MSA Small
150M parameter
Accepts text with/without tashkeel
Voice cloning
Latency: 1.9s (100 characters)
Support English & Context switching

Modern Standard Arabic
SILMA TTS MSA Small
150M parameter
Accepts text with/without tashkeel
Voice cloning
Latency: 1.9s (100 characters)
Support English & Context switching

Our Open-source TTS Models

Our APIs

Integrate Now

API Specifications

Try our Models Now