Pakistan has marked a major milestone in artificial intelligence with the launch of its first homegrown Urdu-language AI model, Qalb. The project has been developed by Taimoor Hassan, a Pakistani student currently studying in the United States.
Qalb is being described as the world’s largest Large Language Model built exclusively for the Urdu language. The model has been trained on 1.97 billion tokens and evaluated using more than seven international benchmarking frameworks. According to its developers, Qalb outperforms existing Urdu-focused AI models on several key real-world performance indicators.
Taimoor Hassan is a young technology entrepreneur who has launched and exited 13 startups. He is a former Microsoft Cup winner and has represented Pakistan on multiple international platforms. He completed his Bachelor’s degree in Computer Science from FAST University and is currently pursuing a Master’s degree in Computer Science and Software Engineering at Auburn University, where he continues to work on advanced AI and technology projects.
With more than 230 million Urdu speakers worldwide, Qalb aims to bridge a long-standing technology gap by enabling advanced AI capabilities in Pakistan’s national language. The model has been designed to support local businesses, startups, educational platforms, digital services, and voice-based AI applications, making technology more accessible to a wider population.
Speaking about the initiative, Hassan said the project is part of a broader mission to contribute to Pakistan’s technological future. He added that the work was carried out in close collaboration with his undergraduate teammates Jawad Ahmed and Muhammad Awais, and that the team plans to continue refining localized AI models tailored to niche industries.
The team emphasized that innovation is no longer limited to large corporations. According to Hassan, modern technology now enables small, focused teams to build powerful tools that can educate, automate, and serve millions of users—without requiring massive budgets or infrastructure.
