WORLDTECH NEWS Global technology intelligence.Contact
← Back to WORLDTECH
AI SINGLE SOURCE

TII Releases Falcon-ASR: 1.6B Arabic Speech Model Hits 20.92% WER

Focused view of a modern data server rack with blinking lights in a blue-lit environment.
Illustrative photo.Photo by panumas nikhomkhai on Pexels

What happened

TII has released Falcon-ASR, a 1.6B speech recognition model for Arabic with an Emirati focus. The Technology Innovation Institute (TII) in Abu Dhabi has released Falcon-ASR , a 1.6B parameter Arabic speech recognition model.

It reports 20.92% average WER on 6 Arabic test sets, but weights and an API (the interface one piece of software uses to talk to another) are not yet available. It focuses on the Emirati dialect and also transcribes English, French, Spanish and Portuguese.

In TII’s evaluation, Falcon-ASR posts a 20.92% average word error rate (WER) across 6 Arabic test sets. For developers building Arabic voice products, the catch is access: today it runs only as a hosted demo.

Key facts

  • TII has β€” released: Falcon-ASR, a 1.6B speech recognition model for Arabic with an Emirati focus
  • The Technology Innovation Institute (TII) in Abu Dhabi has β€” released: Falcon-ASR , a 1.6B parameter Arabic speech recognition model

Sources & evidence