Arabic Dialects · Speech to Text

Jordanian · أردني

Jordanian Arabic
Speech to Text

An STT model built for Jordanian Arabic, understanding the clear Ammani dialect and converting BPO, customer service, and e-commerce calls to accurate transcripts.

dialect: "jordanian"Ammani voice. Real accuracy.

7M+

Native Jordanian Arabic speakers

<15.5%

WER on Jordanian colloquial speech

<300ms

First-word streaming latency

#1 Arabic BPO

Jordan, the world's top Arabic BPO destination

Try it

Speak Jordanian Arabic.
Watch it transcribe.

Record yourself speaking Jordanian 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

Jordanian Arabic Overview

What is Jordanian Arabic?

Jordanian Arabic is a Levantine dialect known for its clarity and moderate phonology, making Amman the world's leading Arabic BPO destination. The Ammani urban variety is the standard in business, contact centers, and digital services.

  • Clear neutral Ammani phonology, the most intelligible Levantine variety.
  • High Arabic-English code-switching in professional settings.
  • BPO and contact center optimized, purpose-built for Jordan's #1 industry.
  • Regional Bedouin and rural varieties covered.

Native speakers

7M+

Jordan

BPO standing

#1 Arab

World's top Arabic BPO destination

Standard variety

Urban Ammani

Clearest and most neutral Levantine

Hamsa WER

<15.5%

Hamsa benchmark 2024

The problem

Why generic Arabic STT fails
on Jordanian Arabic

A model trained on MSA achieves 40–53% word error rate on Jordanian BPO audio. Here is why.

Generic models miss Ammani clarity characteristics

The specific phonological traits that make Ammani Arabic clear and BPO-appropriate are underrepresented in generic training data, causing higher error rates on exactly the variety Jordan's contact centers use.

Arabic-English code-switching in commercial settings

Amman professionals heavily mix Arabic and English, generic models produce inconsistent output when encountering mid-sentence language switches common in Jordan's business environment.

Rural and Bedouin Jordanian varieties underrepresented

Models optimized only for Ammani miss speakers from outside the capital: Jordan's rural and Bedouin populations have distinct phonological features that require dedicated coverage.

BPO environments demand reliability at scale

Contact centers transcribing thousands of calls daily need consistent below-15% WER at scale, generic models degrade under volume and with the telephony noise typical of BPO environments.

The solution

How Hamsa handles
Jordanian Arabic natively

✓

Model trained on real Ammani speaker data

Native Jordanian Ammani speakers across multiple professional environments, call center agents, e-commerce reps, government workers, providing authentic training data.

✓

Arabic-English code-switching support

A parallel detection pass handles intensive professional mixing without accuracy loss, transcribing each segment in its native form.

✓

Coverage beyond Ammani

Data from Jordanian regional varieties, rural and Bedouin communities, to handle the full national linguistic diversity.

✓

BPO-scale reliability

Tested at call center volume and scale for consistent performance, purpose-built for the demands of Jordan's BPO industry.

Capabilities

Everything you need for
Jordanian Arabic STT

✓

Ammani Clarity Recognition

Native Jordanian Ammani speakers across multiple professional environments, capturing the clear neutral phonology that defines the dialect.

✓

Arabic-English Code-Switching

Handles intensive professional mixing without accuracy loss, common in Amman's business and tech environments.

✓

Real-time Streaming Sub-300ms

Sub-300ms first-word latency optimized for BPO contact center call flows.

✓

Speaker Diarization

Identify and label individual speakers in call recordings, critical for Jordan's BPO centers processing thousands of daily calls.

✓

Automatic Punctuation

Sentence boundaries, commas, and question marks inserted correctly for Jordanian Arabic syntax.

✓

Noise Robustness for Contact Centers

Tested at call center volume and scale for consistent performance in noisy telephony environments.

Transcript examples

Real Jordanian Arabic.
Accurate transcription.

BPO call center

شو الخدمة اللي بدكم ياها؟

"What service do you need?"

Clear Ammani dialect captured accurately for BPO

E-commerce

بدي أرجّع الطلب...

"I want to return the order..."

Customer service Jordanian accurately recognized

eGovernment

بدي أجدّد رخصة السيارة...

"I need to renew my car registration..."

Government service terms in Jordanian dialect accurately recognized

Use cases

Where Jordanian Arabic STT
delivers results

BPO & Contact Centers

BPO call automation and analytics

Transcribing and analyzing thousands of daily calls in Jordan's BPO centers serving clients across the Arab world, real-time QA, agent coaching, and compliance monitoring.

E-Commerce

E-commerce customer call transcription

Accurate recognition of customer queries and complaints in Jordanian and regional e-commerce platforms, turning voice interactions into searchable, analyzable data.

Digital Government

Jordan eGovernment services

Accurate Jordanian speech recognition for the eGovernment Jordan platform and unified gateway, enabling voice interfaces that understand how Jordanian citizens actually speak.

Healthcare

Medical consultation transcription

Recognizing Jordanian medical speech in digital health apps and consultation documentation, supporting Jordan's growing digital health sector.

Accuracy benchmark

Jordanian Arabic Speech Recognition Benchmark available

Tested on real Jordanian speech across BPO, e-commerce, and government service environments. Result: under 15.5% WER compared to 40–53% for generic models.

Word Error Rate (WER) · lower is better

Hamsa STT (Jordanian)us15.5%
Generic Arabic model40.2%
MSA-trained model49.8%
GPT-4o-transcribe53.1%

Evaluated on Jordanian 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, whatever your telephony platform or recording system produces.

02

Detect Ammani dialect + code-switching

Hamsa detects Ammani dialect features and routes to the Jordanian-specific acoustic and language models. If code-switching is detected, the parallel Arabic/English decoder activates.

03

Clean text with punctuation + diarization

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 Jordanian Arabic speech recognition

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

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

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

stream.on('transcript', (t) => console.log(t.text));
// → "شو الخدمة اللي بدكم ياها؟" · 00:01.240

Related services

More Hamsa for Jordanian Arabic

Why Hamsa

Built for Arabic from the ground up

✓

Native, not adapted

Hamsa trains on real Jordanian Ammani speaker data, not an English or MSA model fine-tuned for Arabic. The difference is 25–38 WER points on BPO call audio.

✓

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 BPO call rates.

✓

13+ dialects, one API

Switch dialect with a single parameter. No separate integrations, no separate contracts.

FAQ

Jordanian Arabic STT ·
frequently asked

Does Hamsa STT understand the Ammani dialect specifically?

Yes. Hamsa's Jordanian Arabic model is trained on native Ammani speakers in professional environments, the variety used in Jordan's BPO centers, contact centers, and digital services. It captures the clear, moderate phonology that distinguishes Ammani from other Levantine varieties.

Is it suitable for BPO and IVR telephony environments?

Yes. The model delivers sub-300ms first-word latency and is tested at call center volume. It handles telephony noise, multiple speakers, and the rapid speech rates common in BPO environments.

Is it compatible with Jordan eGovernment services?

Yes. The model understands the full range of Jordanian Arabic spoken in government service contexts, from formal requests to colloquial questions, matching how citizens actually interact with digital government platforms.

Does it handle Arabic-English code-switching?

Yes. Amman professionals frequently mix Arabic and English within sentences. Hamsa detects these switches in real time and applies the correct decoder for each segment, transcribing mixed utterances accurately.

Can I add a custom term list for e-commerce or specialized domains?

Yes. A custom vocabulary layer accepts domain-specific terms, product names, brand terms, medical or legal terminology, to improve recognition accuracy in specialized environments.

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.