Uplift AI, a Pakistan-based voice technology company, has developed Urdu-speaking AI voice models now being used by Khan Academy to localize educational content. The move has enabled the global learning platform to publish over 100 Urdu-dubbed videos, with more in the pipeline.
Khan Academy recently used Uplift AI’s Urdu voiceovers to publish 104 educational videos and is currently working on dubbing thousands more. The initiative addresses a long-standing challenge in localizing content for Urdu-speaking audiences.
Until now, AI-generated voice solutions from platforms such as Google, OpenAI, and ElevenLabs had limited success in handling the linguistic nuances of Urdu. This often resulted in high costs and longer turnaround times for content localization.
Founded in 2024 and incubated at Y Combinator, Uplift AI says its models reduce dubbing time by a factor of 10 and significantly lower production costs compared to traditional professional voiceover workflows. These efficiencies could help expand the availability of educational material in regional languages.
The development was announced by Uplift AI CEO Hammad Malik via a LinkedIn post. He also shared that the company will soon roll out new voice models for Balochi and Sindhi.
According to the company’s website, Uplift AI builds speech generation and understanding models tailored to Pakistani languages. The team, which includes former developers from Siri and Alexa, develops and deploys its technology stack in-house and offers access through APIs, SDKs, and web platforms.
This Urdu voiceover project may contribute to broader efforts in making educational and digital services more accessible in regional languages across Pakistan.