Arabic Dialects · Speech to Text

Algerian · جزائري

Algerian Arabic
Speech to Text

An STT model trained specifically on Algerian Arabic, built to understand the deep fusion of Arabic, French and Amazigh as Algerians actually speak it. High accuracy for fintech, telecom, and customer service.

dialect: "algerian"40M+ speakers · real-time · Arabic-French-Amazigh

40M+

Algerian Arabic speakers

<40.5%

WER on Algerian colloquial speech

<300ms

First-word streaming latency

3

Languages interwoven in everyday speech: Arabic, French and Amazigh

Coming Soon

Demo launching soon

The interactive demo for this dialect is coming soon. Sign up to be notified when it launches.

Join the waitlist →

Dialect overview

What is Algerian Arabic?

Algerian Arabic, known as Darja (الدارجة), is a Maghrebi dialect spoken by more than 40 million Algerians and by millions more in France and Europe. It is defined by a deep fusion with French that goes beyond borrowed vocabulary into structure and rhythm, and Amazigh adds its own texture, especially in the Kabyle and Aures regions.

Key features that distinguish Algerian Arabic from Modern Standard Arabic:

  • First-person present starts with n-: "نكتب" means "I write", and "نكتبو" means "we write".
  • Negation wraps the verb in "ما...ش": "ما نعرفش" means "I don't know".
  • French verbs conjugated with Arabic grammar: "نشارجي" (from charger), "نريزرفي" (from réserver).
  • Everyday expressions: "واش راك؟" (how are you?), "بزاف" (a lot), "درك" (now), "يعطيك الصحة" (thank you).

Native speakers

40M+

Algeria

Algerian diaspora

France

Millions in France and Europe

Code-mixing

3

Arabic · French · Amazigh

Hamsa WER

<40.5%

on colloquial Algerian audio · 2024

The problem

Why generic Arabic STT fails
on Algerian Arabic

Algerian speech breaks the assumptions most Arabic models are built on. Here is why.

MSA models do not understand Algerian Darja

Most Arabic STT models are trained on MSA and read-aloud news, while Algerians speak a Maghrebi dialect in daily life that differs from MSA in pronunciation, morphology and vocabulary. Words like "واش" (what), "بزاف" (a lot) and "درك" (now) do not exist in MSA dictionaries, so generic models misread them.

Deep fusion with French

French in Algerian speech is not a handful of borrowed words. It is woven into structure and rhythm: French verbs conjugated with Arabic grammar, and whole sentences that move between the two languages. Generic models treat this as noise or force it into one language, and half the sentence is lost.

The Amazigh layer and regional variety

Amazigh (Tamazight) adds its own texture to Algerian speech, especially in the Kabyle and Aures regions, and pronunciation and vocabulary shift between Algiers, the major cities and the rest of the country. A model that has never heard this variety gets it wrong.

Maghrebi rhythm and dropped vowels

Algerian speech tends to drop short vowels and stack consonants into consecutive clusters, at a fast pace very different from MSA and Levantine or Gulf dialects. Models trained on those varieties read the compressed syllables as errors or unknown words.

The solution

How Hamsa handles
Algerian Arabic natively

✓

Real Algerian training data

Models built on data from real Algerian speakers in Algiers, the major cities and beyond, drawn from calls and everyday conversation, not repurposed MSA data.

✓

A language model designed for Algerian Darja

A language model that knows the vocabulary and structures actually used in Algeria, from "واش راك" (how are you) and "كيفاش" (how) to "ما...ش" negation and French verbs conjugated with Arabic grammar.

✓

A smart code-mixing layer

The model detects switches between Arabic, French and Amazigh automatically within the same conversation, and processes each segment in its own language to produce a transcript that keeps the speech exactly as it was said.

✓

Regional variety coverage

Training covers speakers from Algiers, the major cities and several regions, so the model handles differences in pronunciation and vocabulary without a separate setup for each region.

Capabilities

Everything you need for
Algerian Arabic STT

✓

Native Algerian Transcription

Purpose-trained on Algerian speech as it is actually spoken. Not MSA with a dialect layer on top, but native Algerian acoustic and language models built from the ground up.

✓

Arabic-French-Amazigh Code-Mixing

Algerians move between Arabic, French and Amazigh within the same sentence. Hamsa handles these transitions and writes every word in its own language without losing context.

✓

Speaker Diarization

Identify and label individual speakers in call recordings. Critical for fintech and telecom contact centers processing thousands of calls a day.

✓

Smart Punctuation

Sentence boundaries, commas and question marks placed correctly for the structure of Algerian speech, not MSA grammar rules.

✓

Real-time Streaming

Under 300ms first-word latency. Transcript fragments arrive while the speaker is still talking, with no waiting for the end of the utterance.

✓

Custom Vocabulary

Add Algerian brand names, payment services and industry terms, including French terminology, to improve recognition accuracy in your domain.

Transcript examples

Real Algerian Arabic.
Accurate transcription.

Fintech support

السلام عليكم، حاولت نخلّص بالكارط بصّح العملية ما فاتتش، واش ندير درك؟

"Hello, I tried to pay with my card but the transaction didn't go through. What do I do now?"

Algerian: "بصّح" (but), "درك" (now), "الكارط" (card, from French carte)

Telecom call

الكونكسيون راهي طايحة من البارح، والله ما نقدر نخدم بلا أنترنت.

"The connection has been down since yesterday. Honestly, I can't work without internet."

Natural Algerian: "راهي" (it is), "البارح" (yesterday), "الكونكسيون" (from French connexion)

Diaspora customer

Franchement الخدمة مليحة بزاف، بصّح لازم تريزرفي من قبل، c'est obligatoire.

"Honestly, the service is really good, but you have to book in advance. It's mandatory."

Code-switching: "franchement", "c'est obligatoire", "تريزرفي" (réserver, conjugated in Arabic)

Use cases

Where Algerian Arabic STT
delivers results

Fintech

Digital Payment Call Transcription

As digital payments and mobile wallets expand in Algeria, Hamsa transcribes and analyzes customer calls for compliance and quality monitoring, understanding the French terms Algerians use when they talk about payments and accounts.

Telecom

Customer Service Call Transcription

Algeria's telecom sector is growing alongside rising internet and mobile penetration. Hamsa transcribes customer service calls in real time and helps analyze call drivers to improve service quality.

Algerian Diaspora

Diaspora Services in France

Millions of Algerians in France and Europe speak a blend of Darja and French. Hamsa understands that blend, so businesses can serve the Algerian diaspora in the language it actually speaks, in contact centers and apps.

Digital Government

Digital Citizen Services

Digital government services need voice interfaces that understand Algerian citizens as they really speak, in Darja mixed with French, not in formal MSA alone.

Accuracy benchmark

High-accuracy STT for Algerian Arabic

Measured on real colloquial Algerian speech, from calls and natural conversation mixed with French. Lower word error rate means more words transcribed correctly.

Word Error Rate (WER) · lower is better

Hamsa STT (Algerian)us40.5%
Generic Arabic STT38.4%
MSA-trained model48.1%
GPT-4o-transcribe52.6%

Evaluated on Algerian colloquial 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, so it works with whatever your microphone, telephony platform, or recording system produces.

02

Dialect routing

Hamsa detects Algerian Arabic features automatically and routes audio to the dedicated Algerian acoustic and language models. When code-mixing is detected, Arabic, French and Amazigh segments are each processed in their own language to keep the transcript accurate and the conversation in context.

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 Algerian Arabic speech recognition

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

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

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

stream.on('transcript', (t) => console.log(t.text));
// → "واش راك خويا؟ لاباس عليك؟" · 00:01.840

Related services

More Hamsa for Algerian Arabic

Why Hamsa

Built for Arabic from the ground up

✓

Native, not adapted

Hamsa trains on real Arabic dialect data, not English models adapted for Arabic. The difference is 25 to 50 WER points on colloquial 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 A/B tests across Algerian and Moroccan. No separate integrations, no separate contracts.

FAQ

Algerian Arabic STT,
frequently asked

What is the difference between Algerian Arabic and Modern Standard Arabic?

Modern Standard Arabic is the formal register used in media, education and official documents. Algerian Arabic is the spoken language of more than 40 million Algerians. It is a Maghrebi dialect with its own character, where Arabic is interwoven with French and Amazigh, and it differs from MSA in pronunciation, vocabulary and sentence structure. That is why STT models trained on MSA struggle with natural Algerian speech.

Does Hamsa handle Arabic-French code-mixing in Algerian speech?

Yes. Arabic-French mixing in Algerian speech is a fusion of structure and rhythm, not just borrowed words. Hamsa detects the switch within the same sentence and processes each segment in its own language, and it also handles the Amazigh layer in everyday speech.

What word error rate does Hamsa achieve on Algerian Arabic?

Hamsa achieves a word error rate (WER) under 40.5% on real colloquial Algerian speech. The gap is clearest on informal speech, such as customer service calls and natural conversation mixed with French, which is exactly the data businesses need to process.

Is the model suited to real-time contact centers?

Yes. Real-time streaming with under 300ms first-word latency is designed for contact centers and voice assistants, with speaker diarization for multi-party calls.

Which industries use Algerian Arabic STT most?

The leading use cases are fintech and digital payment services for call transcription and compliance, telecom for transcribing and analyzing customer service calls, businesses serving the Algerian diaspora in France and Europe, and digital government services for citizens.

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.