Arabic voice cloning
Clone an Arabic voice from a few seconds of audio.
Clone any voice from a 4–7 second recording, in Studio or through the API, and use it for text to speech and your AI agents.
Voice clone
Record or upload a WAV, trim a 4–7 second segment, then save it as a workspace voice.
sample.wav· 0:18
0:004–7 s0:18
I have the right to use this voice.
Clone voiceYour voices
concierge_saudi
Cloned voiceNot cloned yet
Ready to use in
- Text to speech in Studio
- Voice agents
- The TTS API
At a glance
- Sample
- A 4–7 second segment from a WAV recording
- Where
- Nabrah Studio or the API
- Use it in
- Text to speech and voice agents
- Control
- Delete a cloned voice at any time
A new voice, without a recording studio.
- 01
Seconds, not hours
A 4–7 second segment of clean speech is enough to create a voice.
- 02
Trim in Studio
Pick the cleanest stretch of your recording on the waveform before cloning.
- 03
One voice everywhere
Use the clone in Studio, the TTS API and your voice agents.
- 04
Clone by API
Create voices with a single multipart request.
- 05
Quick generate
Speak a line in a recorded voice without saving a clone first.
- 06
You stay in control
List your voices and delete any clone at any time.
In Studio
Trim it, name it, clone it.
Studio shows your recording as a waveform. Move the selection over a clean 4–7 second stretch, name the voice, and clone it.
Text to speech
Write your script, tune the model and voice, then render polished audio in one pass.
حياك الله بس حبينا نأكد لك إن الحجز تم تأكيده. وراح نكون بانتظارك في الموعد المحدد لخدمتك بإذن الله.
Voice
Model
Nabrah TTS v1
Voice clone
Record or upload a WAV, trim a 4–7 second segment, then save it as a workspace voice.
0:004–7 s0:18
Everywhere you use voice
One voice, every product.
A cloned voice sits next to the public voices. Use it for scripts in Studio, as the voice in TTS requests, or as the voice your agent speaks with on calls.
Voices
The public library and the voices you clone.
Search voices
Your voices
Public voices
API
Clone from code.
Send the recording and a name to the clone endpoint. Check the voice until it is ready, then pass its id as the voice in TTS or as an agent's voice.
curl -X POST https://api.nabrah.ai/api/ext/tts/clone_voice \
-H "X-API-Key: $NABRAH_API_KEY" \
-F "name=concierge_saudi" \
-F "audio_file=@sample.wav"{ "success": true, "voice_id": "…" }Trimmed for clarity. Every field is in the API reference.
How it works
Your voice in three steps.
Use a clean recording of a voice you have the right to use.
- 1
Record or upload
Record in Studio or upload a clean WAV of the voice.
- 2
Trim and name it
Pick a 4–7 second segment of clear speech and give the voice a name.
- 3
Use it everywhere
When the voice is ready it appears in your voices for Studio, the API and your agents.
Where teams use cloned voices.
One recognisable voice across every place customers hear you.
Give your phone line, ads and app one consistent voice.
Daniah
Brand voice
مرحبا بكم في نبرة، صوت واحد في كل مكان.
Frequently asked questions
Straight answers about Nabrah Voice Cloning. Anything else, our team is a message away.
Voice cloning creates a synthetic voice that sounds like a real speaker from a sample of their speech. In Nabrah, a cloned voice can read any Arabic text in Studio, through the TTS API, or as the voice of an AI agent.
A clean WAV recording. In Studio you select a 4–7 second segment of clear speech from it; the API takes the WAV file directly.
Only clone a voice you own or have clear permission to use. Voice cloning on Nabrah is meant for your own and your brand's voices.
Everywhere you use Nabrah voices: Studio scripts, TTS API requests, and your voice agents.
Yes. Remove it from your voices in Studio or with the delete-voice endpoint in the API.
Cloning a voice uses 250 credits from your plan. Speech generated with it is priced like any text to speech.
One platform. Every voice product.
Nabrah Agents
AI voice agents for calls, WhatsApp and your website.
ExploreAppointment HandlingLead QualificationNabrah Text to Speech
Natural Saudi voices from text, in Studio or by API.
ExploreNabrah Speech to Text
Saudi and Gulf Arabic transcription, live or from files.
Exploreالسلام عليكم. حاب أحجز موعد بكرة