---
title: "Arabic dialect speech recognition - how to test ASR | BroadRouter"
description: "Why speech-to-text struggles with Gulf, Egyptian, Levantine and Maghrebi Arabic, and how to build a dialect test set that shows how a provider will do on your audio."
url: https://broadrouter.com/blog/arabic-dialect-speech-recognition/
lang: en
updated: 2026-11-14
publisher: BroadRouter (a BroadNet site); API by BroadNet Inference, https://inference.broadnet.ai
---

# Arabic dialect speech recognition: Gulf, Egyptian, Levantine, Maghrebi

By the BroadNet Inference team · 5 min read · Published 2026-11-14 · Updated 2026-11-16

Arabic dialects differ from Modern Standard Arabic in sounds, words and spelling, and speakers mix in English or French, so ASR models trained mostly on formal Arabic score far worse on them. To test a provider, build a test set per dialect from your real audio, fix a spelling convention, and score WER, CER and entity accuracy per dialect.

## Key takeaways

-   Dialect is a major accuracy driver in Arabic ASR: BroadNet reports 3.8 WER on read Arabic and 23.4 on dialect speech for the same model.
-   Dialects have no standard spelling, so decide whether references are verbatim dialect or normalised before you score.
-   Gulf speech mixes in English; Maghrebi speech mixes in French. Score code-switched clips as their own group.
-   Report WER and CER per dialect, plus accuracy on numbers and names. Never trust one averaged number.
-   Thirty to fifty clips per dialect, with ten hand-checked references, is enough to rank providers.

**On this page**

1.  Why are Arabic dialects hard for speech recognition?
2.  How does each Arabic dialect trip up speech recognition?
3.  How do you build an Arabic dialect test set?
4.  What do published Arabic WER numbers tell you?
5.  Which LLM should read dialect transcripts?
6.  Where BroadNet Inference fits

Speech-to-text struggles with Arabic dialects because they differ from Modern Standard Arabic (MSA) in sounds, words and spelling, and speakers mix in English or French. To know how a provider will do, build a small test set per dialect from your real audio and score each dialect separately.

A Riyadh caller, a Cairo voice note and a Casablanca support call are all “Arabic”, and one model can handle the first well and the third badly.

This post explains what makes each major dialect group hard for speech recognition, and how to build a test that shows how a provider will do on **your** dialects. For the broader buying checklist (price, limits, privacy), see the [Arabic speech-to-text buyer’s guide](https://broadrouter.com/blog/arabic-speech-to-text-api-buyers-guide/).

## Why are Arabic dialects hard for speech recognition?

Four problems stack up:

1.  **Different sounds.** The same letter is pronounced differently by region, so the acoustic model must map several sounds to one spelling.
2.  **Different words.** Everyday words differ by dialect. A model trained mostly on MSA may “correct” a dialect word into its formal equivalent, which counts as an error if your reference is verbatim.
3.  **No standard spelling.** Dialects are written informally, so even two human transcribers disagree. Training data is uneven too: a 2024 study of dialectal coverage notes that existing ASR systems “mainly cover the modern standard Arabic (MSA) variety and few high-resource dialects” ([arXiv 2411.05872](https://arxiv.org/abs/2411.05872)).
4.  **[Code-switching](https://broadrouter.com/glossary/code-switching/).** Speakers mix English or French mid-sentence. The [Casablanca](https://aclanthology.org/2024.emnlp-main.1211) dataset, which covers eight dialects, annotates code-switching alongside dialect and gender.

The [Open Universal Arabic ASR Leaderboard](https://arxiv.org/abs/2412.13788) paper (Interspeech 2025) evaluates open-source models across several multi-dialect test sets for exactly this reason: one Arabic score hides the spread.

## How does each Arabic dialect trip up speech recognition?

| Dialect group | Where | Typical ASR trouble spots |
| --- | --- | --- |
| Gulf (Khaleeji) | Saudi Arabia, UAE, Kuwait, Qatar, Bahrain, Oman | English product names, numbers and brand terms mixed in; regional pronunciation of qaf and kaf |
| Egyptian | Egypt | Jeem pronounced as a hard g; qaf often a glottal stop; very common in media, so models usually see more of it |
| Levantine | Syria, Lebanon, Jordan, Palestine | Urban qaf as a glottal stop; French or English mixed in, especially in Lebanon |
| Maghrebi (Darija) | Morocco, Algeria, Tunisia, Libya | Furthest from MSA; short vowels dropped; heavy French code-switching; Amazigh loanwords |
| Iraqi | Iraq | Distinct vocabulary and pronunciation; less public training audio |

These are tendencies, not rules. City and rural speech differ inside each group, and younger speakers code-switch more. The point is that each group fails in its own way, so you score each group on its own.

## How do you build an Arabic dialect test set?

### 1\. Sample real audio per dialect

Pull 30 to 50 clips per dialect from the channel you will run in production: phone calls, WhatsApp voice notes, meeting recordings. Phone audio at 8 kHz behaves differently from studio audio. Label each clip by dialect, and tag clips that contain English or French.

### 2\. Agree a spelling convention first

Before anyone writes a reference, decide:

-   **Verbatim dialect** (write what was said, in dialect spelling), or
-   **Normalised** (map to a consistent form).

Write the choice down with examples. If references are verbatim and the model outputs MSA-like words, WER will look terrible even when the meaning is right. Neither choice is wrong. It has to match what your downstream system needs.

### 3\. Write references for 10 clips per dialect

Have native speakers of that dialect transcribe them. For code-switched clips, decide whether English words stay in Latin script, and apply it everywhere.

### 4\. Normalise both sides, then score

Strip diacritics, unify alef forms (إ أ آ to ا) and final yaa (ى to ي), remove punctuation, and convert Eastern Arabic digits to Western ones. Then compute:

| Metric | Why |
| --- | --- |
| [WER](https://broadrouter.com/glossary/word-error-rate/) per dialect | The standard measure; strict on clitics and prefixes |
| CER per dialect | More forgiving of spelling variation |
| Entity accuracy | Share of numbers, dates and names transcribed exactly |
| Code-switch group WER | Shows whether English or French words survive |

The `jiwer` package computes WER and CER; [Arabic WER explained](https://broadrouter.com/blog/arabic-asr-wer-explained/) has a normalisation function you can reuse and a worked example of how much it changes the score.

### 5\. Read the transcripts

Numbers decide whether a call-centre or banking workflow works. Read 20 transcripts per dialect and check every number, date and name by hand.

## What do published Arabic WER numbers tell you?

Public benchmarks give a direction, not a promise. BroadNet publishes three WER figures for BroadNet Transcribe Arabic in its [model catalogue](https://www.broadnet.ai/inference/models): 3.8 on FLEURS-ar (clean read speech), 21.1 on MASC-ar (mixed real-world speech) and 23.4 on its own dialect set. That is a sixfold jump for the same model between read speech and dialect speech, which is why you test on your own dialects.

When you compare providers, compare them on the same test set. A leaderboard such as the [Open Universal Arabic ASR Leaderboard](https://huggingface.co/spaces/elmresearchcenter/open_universal_arabic_asr_leaderboard) is useful because every model is scored the same way. A vendor’s own number on its own set is not comparable with another vendor’s.

## Which LLM should read dialect transcripts?

Transcripts usually feed an LLM: to summarise a call, route a ticket or answer a question. That model needs to read dialect too. Our [guide to choosing an LLM for Arabic](https://broadrouter.com/blog/choosing-an-llm-for-arabic/) covers how to test it, and [Qwen3.8 27B](https://broadrouter.com/models/qwen3.8-27b/) is the model BroadNet tested on Gulf and Egyptian dialects.

## Where BroadNet Inference fits

[BroadNet Transcribe Arabic](https://broadrouter.com/models/cohere-transcribe-arabic/) is built for Arabic including dialects, costs $0.0042 a minute ($0.25 an hour), and BroadNet reports it #2 of 37 (its September 2026 figure; the live board listed it #4 of 39 on 15 November 2026) on the Open Universal Arabic ASR Leaderboard, and #1 among models whose licences permit serving through an API. For numbers, dates and names in noisy audio, BroadNet offers a fine-tuned version [on request](https://broadrouter.com/models/#on-request). The $5 starting credit covers about 1,190 minutes, enough for the full test above; the [speech-to-text cost calculator](https://broadrouter.com/tools/speech-to-text-cost-calculator/) turns your production volume into a monthly bill. [Create a key](https://inference.broadnet.ai/?utm_source=broadrouter&utm_medium=referral&utm_campaign=blog-arabic-dialect-speech-recognition&utm_content=body) to start.

BroadNet does not publish separate WER figures for each dialect group. That is exactly why you should run this test: its dialect-set figure is one number, and your customers speak one or two dialects, not all of them.

Free tools for this

-   [Arabic token counter](https://broadrouter.com/tools/arabic-token-counter/) Arabic vs English token counts per tokenizer: the "Arabic tax" in tokens and money.
-   [Arabic text normalizer](https://broadrouter.com/tools/arabic-text-normalizer/) Tashkeel, tatweel, alef, ya, ta marbuta, digits and invisible characters, with a matching Python function.
-   [Speech-to-text cost calculator](https://broadrouter.com/tools/speech-to-text-cost-calculator/) Hours, calls or voice notes to a monthly transcription bill, against your current provider.

[All free tools](https://broadrouter.com/tools/)

## Terms in this article

[Speech-to-text](https://broadrouter.com/glossary/speech-to-text/)

Speech-to-text, also called automatic speech recognition (ASR), is technology that converts spoken audio into written text. A speech-to-text API accepts an audio file or stream and returns a transcript, usually priced per minute of audio. Accuracy is measured with word error rate and varies with accent, dialect and recording quality.

[Word error rate](https://broadrouter.com/glossary/word-error-rate/)

Word error rate (WER) is the standard accuracy measure for speech recognition: the number of word substitutions, deletions and insertions needed to turn a transcript into the correct reference, divided by the number of words in the reference. Lower is better.

[Dialectal Arabic](https://broadrouter.com/glossary/dialectal-arabic/)

Dialectal Arabic is the everyday spoken Arabic of a region or city, such as Egyptian, Gulf, Levantine or Maghrebi Arabic. Dialects differ from Modern Standard Arabic in sounds, words and grammar, have no standard spelling, and are increasingly written online, which makes them harder for language models.

[Modern Standard Arabic (MSA)](https://broadrouter.com/glossary/msa/)

Modern Standard Arabic (MSA) is the standardized form of Arabic used in writing, news, education and formal speech across the Arab world. It is nobody's native everyday language: people speak regional dialects such as Egyptian or Gulf Arabic and switch to MSA in formal settings.

[Code-switching](https://broadrouter.com/glossary/code-switching/)

Code-switching is the use of two or more languages, or varieties of one language, within the same conversation or sentence, in speech or writing. Examples include Arabic speakers mixing English words into dialect, or moving between Modern Standard Arabic and their dialect.

[Character error rate (CER)](https://broadrouter.com/glossary/character-error-rate/)

Character error rate (CER) measures transcription or OCR accuracy at the character level: the substitutions, deletions and insertions needed to turn the output into the reference text, divided by the number of characters in the reference. Lower is better, and 0 is a perfect match.

[Arabic diacritics (tashkeel)](https://broadrouter.com/glossary/arabic-diacritics/)

Arabic diacritics, or tashkeel, are small marks written above or below letters to show short vowels, doubled consonants and the absence of a vowel. They are optional: most Arabic text, including news, omits them, so readers and software infer the vowels from context.

[Tool calling](https://broadrouter.com/glossary/tool-calling/)

Tool calling, also called function calling, is an LLM API feature in which the model returns a structured request to run a function the developer has defined, with JSON arguments. The application runs the function and sends the result back; the model never runs code itself.

## Questions

### Which Arabic dialect is hardest for speech recognition?

There is no universal ranking; it depends on the model's training data. Maghrebi dialects such as Moroccan and Algerian are often hard because they sit furthest from Modern Standard Arabic and mix in French. Test your own dialect rather than relying on a general ranking.

### Can speech-to-text understand Egyptian Arabic?

Often better than rarer dialects, because Egyptian is widely used in media, but expect lower accuracy than on formal Arabic. Measure it on your own recordings.

### Should I use WER or CER for Arabic?

Use both. WER counts whole words, so a wrong prefix or clitic makes the whole word wrong. CER counts characters and is more forgiving of spelling variation, which matters for dialects with no standard spelling.

### How many audio clips do I need to test an Arabic ASR provider?

About 30 to 50 real clips per dialect, with at least 10 per dialect transcribed by hand as references. That is enough to see large differences between providers.

### Does the language code matter for Arabic dialects?

Send language=ar. Dialects share the ar code on most APIs. Without it, some models return Arabic speech transliterated into Latin letters.

## Keep reading

-   [Arabic speech-to-text API: how to choose one in 2026](https://broadrouter.com/blog/arabic-speech-to-text-api-buyers-guide/)
-   [Choosing an LLM API for Arabic: what to test](https://broadrouter.com/blog/choosing-an-llm-for-arabic/)
-   [Whisper API alternative: when to switch, and how to test one](https://broadrouter.com/blog/whisper-api-alternative/)
-   [How to transcribe call center recordings: an Arabic-ready pipeline](https://broadrouter.com/blog/transcribe-call-center-recordings/)
-   [Translate Arabic audio to English text with an API: a tested pipeline](https://broadrouter.com/blog/translate-arabic-audio-to-english/)
-   [Add diacritics to Arabic text with an LLM: prompt, checks, DER](https://broadrouter.com/blog/arabic-diacritization-llm/)
-   [Arabic grammar and spelling correction with an LLM: a how-to](https://broadrouter.com/blog/arabic-grammar-spelling-correction-llm/)
-   [Guide: Arabic speech-to-text API, step by step](https://broadrouter.com/guides/arabic-speech-to-text-api/)

Try it on your real workload. New accounts start with $5 of credit for Gemma and transcription. DeepSeek, GLM and Qwen3.8 need a top-up.

[Get started](https://inference.broadnet.ai/account/?utm_source=broadrouter&utm_medium=referral&utm_campaign=blog-arabic-dialect-speech-recognition&utm_content=page-cta) ·

**Sources**

-   [Talafha et al., Casablanca: Data and Models for Multidialectal Arabic Speech Recognition (EMNLP 2024; fetched 2026-11-14)](https://aclanthology.org/2024.emnlp-main.1211)
-   [Wang, Alhmoud and Alqurishi, Open Universal Arabic ASR Leaderboard (Interspeech 2025; arXiv 2412.13788; fetched 2026-11-14)](https://arxiv.org/abs/2412.13788)
-   [Dialectal Coverage and Generalization in Arabic Speech Recognition (arXiv 2411.05872; fetched 2026-11-14)](https://arxiv.org/abs/2411.05872)
-   [Open Universal Arabic ASR Leaderboard, Hugging Face Space (fetched 2026-11-14)](https://huggingface.co/spaces/elmresearchcenter/open_universal_arabic_asr_leaderboard)
-   [BroadNet model catalogue, speech to text: WER as reported by BroadNet (fetched 2026-11-14)](https://www.broadnet.ai/inference/models)
