Pocket FM Hits $500M Revenue Run Rate on Back of AI Audio Pipeline

Pocket FM has reached an annualized revenue run rate of $500 million, doubling its commercial scale over the past twelve months by leaning heavily on artificial intelligence to build out its serialized audio library.

Automated systems now generate 93 percent of the platform's total catalog and 99 percent of its new releases, according to Rohan Nayak, co-founder and chief executive of Pocket FM. In an interview with TechCrunch, Nayak emphasized that human creators continue to supply foundational ideas and plotlines, while machine learning systems execute the writing and voice output. "We want to create great IPs that last 100 years, and that needs humans," Nayak said.

Vasu Sharma, head of AI at Pocket FM and a former researcher at Meta and Tesla, noted that the startup trained proprietary models on years of historical listener engagement data to manage creative writing and text-to-speech synthesis. The transition has made content creation roughly 80 times cheaper. One hundred hours of audio, which previously required a year to produce, can now be generated in a single day. A base of over 550,000 creators now produces 2.5 million hours of AI-assisted content annually across a library exceeding 770,000 series.

That production volume has translated directly into higher subscriber retention. Pocket FM said its 12-month revenue retention rate improved to 76 percent from 44 percent two years ago. The business model relies primarily on micropayments, with roughly $415 million generated from listeners unlocking individual episodes and $85 million from advertising. The United States has emerged as its top market, accounting for 70 percent of annualized revenue after growing 70 percent over the past year. To date, 96 titles have generated more than $1 million in revenue, including 13 that have exceeded $10 million.

Parent company Pocket Entertainment is also translating the model into short-form video. Its microdrama platform, Pocket Saga, has achieved a $15 million run rate within three months of launch in the U.S. by adapting popular audio intellectual property into fully AI-generated video without traditional live-action filming. The company intends to expand into two additional media formats over the next five years.

The startup's trajectory demonstrates how generative tooling is altering entertainment economics at the margin. By substituting voice actors and production studios with customized synthesis models, digital fiction platforms can test thousands of narrative hooks against user payment behavior at minimal cost, establishing proven entertainment franchises before traditional media companies can finish development cycles.

Pocket Entertainment is now in talks with investors to secure between $100 million and $120 million at a valuation approaching $2 billion. Nayak stated the business is profitable on an adjusted basis and faces no pressure to raise, while ruling out any initial public offering in India or the U.S. for at least the next two years.