← All stories
● Covered by 1 source · 2 reportsLow impact2 neutral

Falcon-Emirati-7B LLM Released for Emirati Arabic Dialect Understanding

🔄 Updated 44d ago — new reporting from Hugging Face Blog
New to BrevFeed? We gather this story from every outlet covering it into one summary — ranked by real-world impact, not just the latest headline — so you never miss what matters. What is BrevFeed? →

Key points

  • Falcon-Emirati-7B specializes in Emirati Arabic dialect.
  • It is built on the Falcon-H1-Arabic 7B model.
  • The model aims to capture cultural context and nuance.
  • It uses a hybrid Mamba and Transformer architecture.
  • Falcon-ASR is a 1.6 billion parameter speech recognition model.
  • It supports English, French, Spanish, and Portuguese in addition to Arabic.
  • Falcon-ASR achieved a 20.92% average word error rate across six Arabic test sets.
  • It outperformed previous best published results by 2.25 percentage points.
  • The model supports word-level timestamps for transcriptions.
  • Falcon-ASR was developed at the Technology Innovation Institute (TII).

Addressing Dialectal Nuance in LLMs

Falcon-Emirati-7B is a new large language model developed to specifically address the intricacies of Emirati Arabic. While Modern Standard Arabic (MSA) is common in formal texts, daily communication in the UAE occurs in Emirati Arabic, a Gulf dialect with distinct vocabulary, rhythm, and cultural context. Existing models often miss these nuances, even if they can translate individual words.

Foundation on Falcon-H1-Arabic

The Falcon-Emirati-7B model is built upon the Falcon-H1-Arabic family, specifically the 7B parameter variant. The Falcon-H1-Arabic models utilize a hybrid architecture combining State Space Models (Mamba) and Transformer attention, which provides efficiency for long sequences and precision for long-range dependencies in morphologically rich languages like Arabic. The base models were already trained on a mix of MSA and various Arabic dialects, along with English and multilingual data.

Specialization for Emirati Arabic

This new model takes the foundation of Falcon-H1-Arabic and further specializes it for the Emirati dialect. This includes adapting to its specific vocabulary, grammar, and cultural knowledge that a general Arabic model might not capture. The 7B parameter size was chosen as a balance between retaining dialectal nuance and maintaining practical training and inference costs.

Updates

🕒 2026-10-08 · new reporting from Hugging Face Blog
  • Falcon-ASR is a 1.6 billion parameter speech recognition model.
  • It supports English, French, Spanish, and Portuguese in addition to Arabic.
  • Falcon-ASR achieved a 20.92% average word error rate across six Arabic test sets.
  • It outperformed previous best published results by 2.25 percentage points.
  • The model supports word-level timestamps for transcriptions.
  • Falcon-ASR was developed at the Technology Innovation Institute (TII).

✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →

The daily brief

One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.

One email a day. Unsubscribe in one click, any time.

Today's brief

Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.

~4 min · 3 stories · Nov 21

▶ Play today's brief Listen on Spotify

New every morning, and the back catalogue is archived by date.

Primary sources

arXiv 2509.07526

How outlets covered it

The Technology Innovation Institute (TII) has released Falcon-ASR, a 1.6 billion parameter speech recognition model primarily for Arabic, including the Emirati dialect, and also supporting English, French, Spanish, and Portuguese. Falcon-ASR achieved a 20.92% average word error rate across six Arabic test sets, outperforming previous best published results by 2.25 percentage points, and recorded the lowest word and character error rates in internal Emirati evaluations.

A new large language model, Falcon-Emirati-7B, has been released, specifically designed to understand and generate Emirati Arabic, including its cultural nuances. This model builds upon the Falcon-H1-Arabic family to address the gap in LLMs for specific Arabic dialects beyond Modern Standard Arabic.