Arabic Dialects · Speech to Text

Palestinian · فلسطيني

Palestinian Arabic
Speech to Text

An STT model that understands Palestinian Arabic in its urban and rural varieties. Built for development organizations, diaspora communities, and media production on every continent.

dialect: "palestinian"5M+ native speakers · real-time · urban & rural

5M+

Native Palestinian Arabic speakers

<16.3%

WER on Palestinian Arabic

<300ms

First-word streaming latency

Urban + Rural

Palestinian variety coverage

Try it

Speak Palestinian Arabic.
Watch it transcribe.

Record yourself speaking Palestinian Arabic and watch the words appear as you talk. The demo runs Hamsa's full production STT, the same model your API calls will use.

Speech to Text · speech in → text out

No sign-up
Microphone
press to record

Your words will appear here.

Press record and speak Arabic or English. The words appear as you talk.

Free to start · in your browserTranscribe in Hamsa Media

Dialect overview

What is Palestinian Arabic?

Palestinian Arabic is a Levantine dialect positioned between Jordanian and Syrian in the dialect continuum, spoken by more than 5 million people in Palestine and with a global presence through diaspora communities on every continent. It spans urban varieties (Jerusalem, Ramallah, Nablus) and rural and village varieties with distinct linguistic features.

What a Palestinian Arabic STT model has to cover:

  • Urban varieties with a modern, cosmopolitan character.
  • Rural and village varieties with traditional linguistic features.
  • Diaspora speech from the Americas, Europe, and the Gulf.
  • NGO and development sector meetings and consultations.

Native speakers

5M+

Palestine

Diaspora

Global

On every continent

Varieties

Urban + Rural

Jerusalem, Ramallah, Nablus and rural

Hamsa WER

<16.3%

on Palestinian audio · 2024

The problem

Why generic Arabic STT fails
on Palestinian Arabic

Generic and MSA-trained models reach 41 to 55% word error rate on Palestinian speech. Here is why.

Urban-rural divergence breaks generic models

Rural Palestinian speakers differ significantly from urban ones, and a non-specialized model fails with both. Pronunciation and vocabulary variation between Jerusalem, Nablus, and village areas requires dedicated training.

Diaspora speech carries layered influences

The Palestinian diaspora in Latin America, Sweden, and the Gulf speaks Palestinian Arabic with different linguistic influences. Generic models cannot handle this variation accurately.

The NGO sector demands high transcription accuracy

Humanitarian and development meetings require reliable transcription. A 40% error rate means untrustworthy documentation and lost information that development work depends on.

Palestinian media deserves professional transcription

Palestinian media content and podcasts need accurate transcripts for archiving, translation, and search. Generic models deliver accuracy too low for professional production.

The solution

How Hamsa handles
Palestinian Arabic natively

✓

Diverse Palestinian training corpus

Training data from speakers in multiple cities and rural areas, including Jerusalem, Ramallah, and Nablus. Geographic diversity in the data ensures comprehensive coverage.

✓

Diaspora speech coverage

Data from Palestinian communities in the Americas, Europe, and the Gulf, so transcription stays accurate for Palestinian communities wherever they are.

✓

Accuracy suited to humanitarian work

Under 16.3% WER is suited to documenting NGO meetings and humanitarian consultations, producing reliable, searchable records for reporting and follow-up.

✓

Media production support

Accuracy suited to transcribing Palestinian broadcasts and podcasts, with support for searchable archives and subtitle generation.

Capabilities

Everything you need for
Palestinian Arabic STT

✓

Urban & Rural Coverage

Trained on Palestinian speech from multiple cities and rural areas, including Jerusalem, Ramallah, Nablus, and village varieties, so it handles the full linguistic spectrum.

✓

Diaspora Speech

Training data includes Palestinian speech from Latin America, Europe, and the Gulf, for accurate transcription of diaspora communities worldwide.

✓

Speaker Diarization

Identify and label individual speakers in meeting and interview recordings. Critical for NGO documentation and media production.

✓

Smart Punctuation

Automatic sentence boundaries, commas, and question marks inserted correctly for Palestinian Arabic syntax, not MSA grammar rules.

✓

Real-time Streaming

Under 300ms first-word latency. Transcript fragments arrive as the speaker talks. No buffering, no silence-detection delay.

✓

Code-Switch Detection

Detect and handle code-switching between Palestinian Arabic and English, transcribing each segment accurately.

Transcript examples

Real Palestinian Arabic.
Accurate transcription.

NGO meeting

بالنسبة للمشروع، اتفقنا إنه نبلّش التوزيع الأسبوع الجاي بالقرى الشمالية.

"As for the project, we agreed to start distribution next week in the northern villages."

Palestinian: "بلّش" (start), "الجاي" (next), "إنه" (that)

Diaspora service call

بدّي أبعت مصاري لأهلي بنابلس، قدّيش بتاخدوا عمولة؟

"I want to send money to my family in Nablus. How much commission do you take?"

Natural Palestinian: "بدّي" (I want), "مصاري" (money), "قدّيش" (how much)

Podcast

هاد الإشي صار trend عند الشباب، كل واحد بدّه يعمل podcast خاص فيه.

"This has become a trend among young people. Everyone wants to make their own podcast."

Code-switching: "trend" and "podcast" mid-sentence, "هاد الإشي" (this thing)

Use cases

Where Palestinian Arabic STT
delivers results

NGO & Development

Development project meeting transcription

Converting meetings and consultations of humanitarian and development organizations in Palestine to accurate text.

Diaspora Services

Palestinian diaspora applications

Accurate transcription for financial transfer and diaspora service calls for Palestinian communities in the Americas and Europe.

Education

Palestinian education platforms

Accurate Palestinian speech recognition for e-learning platforms and virtual schools.

Media

Palestinian media content transcription

Producing accurate transcripts and subtitles for Palestinian media content and podcasts.

Accuracy benchmark

The most accurate Palestinian Arabic STT available

Measured on a held-out benchmark of Palestinian speech from media, diaspora, and development settings. Lower word error rate means more words transcribed correctly.

Word Error Rate (WER) · lower is better

Hamsa STT (Palestinian)us16.3%
Generic Arabic STT40.8%
MSA-trained model50.1%
GPT-4o-transcribe54.7%

Evaluated on Palestinian Arabic speech benchmarks · 2024 · Lower is better

Architecture

How it works

01

Audio in

Send audio via WebSocket stream or REST upload. Accepts PCM, MP3, M4A, WAV, and WebM, whatever your microphone, telephony platform, or recording system produces.

02

Dialect routing

Hamsa detects Palestinian Arabic features and routes to the Palestinian-specific acoustic and language models. If code-switching is detected, Arabic and English segments are decoded in parallel.

03

Transcript out

Transcripts stream as JSON events with word-level timestamps, speaker labels, confidence scores, and punctuation. Pipe directly into your database, UI, or downstream NLP pipeline.

Developer API

Two lines to Palestinian Arabic speech recognition

Drop in the Hamsa SDK, set dialect: "palestinian", and start streaming transcripts. No preprocessing pipeline, no normalization to MSA, no fallback surprises.

Read the API docs →
palestinian-stt.tsstreaming
import { Hamsa } from '@hamsa-ai/sdk';

const stream = hamsa.stt.stream({
  model:       'hamsa-stt-s4',
  dialect:     'palestinian',
  diarization: true,
});

stream.on('transcript', (t) => console.log(t.text));
// → "بدّي أبعت مصاري لأهلي بنابلس" · 00:01.840

Related services

More Hamsa for Palestinian Arabic

Why Hamsa

Built for Arabic from the ground up

✓

Native, not adapted

Hamsa trains on real Palestinian speech from Palestine and the diaspora, not English models adapted for Arabic. The difference shows most on colloquial, unscripted speech.

✓

Enterprise-grade SLA

99.9% uptime, SOC 2 Type II compliance, and data residency in the MENA region. Your audio stays in-region.

✓

Fastest Arabic streaming

Under 300ms first-word latency in production. Real-time transcription that keeps pace with human speech rates.

✓

13+ dialects, one API

Switch dialect with a single parameter. Run Palestinian alongside Jordanian, Syrian, or Gulf. No separate integrations, no separate contracts.

FAQ

Palestinian Arabic STT,
frequently asked

Does it cover urban and rural varieties?

Yes. Training data spans multiple Palestinian cities and rural areas, including Jerusalem, Ramallah, Nablus, and village varieties, for accurate transcription regardless of the speaker's background.

Is it suitable for NGO documentation?

Yes. Under 16.3% WER ensures reliable documentation of meetings and humanitarian consultations, producing records that can be trusted for reporting and decision-making.

Does it work with diaspora speech?

Yes. Training data includes Palestinian speech from the Americas, Europe, and the Gulf, for accurate transcription of Palestinian diaspora communities everywhere.

Does it support English code-switching?

Yes. The model automatically detects code-switching between Palestinian Arabic and English and processes each segment in its own language, for accurate mixed-language transcription.

Can it be customized with humanitarian terminology?

Yes. Custom vocabulary from the relief, development, and humanitarian sectors can be added to improve recognition accuracy in humanitarian work contexts.

Explore more

13+ Arabic dialects, supported

Partner ecosystem

Trusted Across the
Technology Ecosystem

Hamsa works with leading technology companies, integrators, and solution providers to bring enterprise Voice AI to organizations across the region.