What is it about?
Qalb is an AI model built specifically to understand and communicate in Urdu. While over 230 million people speak Urdu, most AI models still struggle with the language. We trained Qalb on 1.97 billion tokens of Urdu and English content, including news, literature, government documents, and social media. Qalb achieved a 90.34 score on Urdu benchmarks, outperforming the previous leading Urdu model and the original LLaMA 3.1 model by a significant margin. In simple terms: Qalb makes AI significantly better at understanding, reasoning, and communicating in Urdu.
Featured Image
Photo by Aerps.com on Unsplash
Why is it important?
It is important because Urdu is spoken by over 230 million people, yet it remains underserved by modern AI. Qalb helps close this gap by making AI better at understanding and communicating in Urdu, giving millions of Urdu speakers access to more capable AI technology.
Perspectives
My perspective is that Qalb can become a foundation for hundreds of new Urdu-language use cases, much like what happened with English when ChatGPT brought advanced AI to the mainstream. It can open the door to new applications in education, content creation, businesses, government, customer service, and everyday communication, helping bring Urdu into the next generation of AI.
Muhammad Taimoor Hassan
Auburn University
Read the Original
This page is a summary of: Qalb: Largest State-of-the-Art Urdu Large Language Model for 230M Speakers with Systematic Continued Pre-Training, ACM Transactions on Asian and Low-Resource Language Information Processing, August 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3839365.
You can read the full text:
Contributors
The following have contributed to this page
