GPT-4o sparks record revenue spike for ChatGPT mobile

Giving away your best product for free sounds like a terrible way to make money. Unless, apparently, you are OpenAI.

On May 13, the company rolled out GPT-4o, its new flagship model packing real-time multimodal smarts, human-like voice conversations, and video reasoning. While the headline announcement promised free access on the web, mobile users were met with a very different reality during the rollout week.

And that little gap just triggered the biggest mobile revenue spike in ChatGPT history.

A masterclass in rollout timing

According to data from app intelligence firm Appfigures, net revenue for the ChatGPT mobile app jumped 22% on the day GPT-4o went live. It only accelerated from there.

By Tuesday, daily net revenue hit $900,000, roughly double the app's daily average of $491,000. These figures, keep in mind, are net earnings after Apple and Google take their app store cuts.

Across the five-day window from May 13 to May 17, the mobile app generated a massive $4.2 million in net revenue across iOS and Android. For comparison, ChatGPT's second-biggest spike back in April looked like a minor bump next to this wave.

So why did mobile users open their wallets for a model that was supposed to be free? Because OpenAI didn't roll out GPT-4o to free mobile users right away. If you opened your phone and wanted to talk to the new model in the first week, you were nudged straight into the $19.99 monthly ChatGPT Plus subscription.

Apple users drove the vast majority of the cash, accounting for 81% of net revenue. Geographic demand was led by the U.S. at $1.8 million, followed by Germany at $282,000, the UK at $212,000, and Japan at $210,000. Other top contributors included France, Canada, South Korea, Brazil, Australia, and Turkey.

Why it matters

We think this spike proves something fundamental: consumer appetite for cutting-edge AI is massive, even when the monthly price tag sits well above standard entertainment subscriptions like Netflix.

From a business perspective, OpenAI played its rollout brilliantly. It scored universal praise by framing GPT-4o as a free gift to the masses, but held back the mobile rollout just enough to trigger a wave of FOMO upgrades. Users who wanted to test the voice capabilities where they mattered most, directly on their phones, were more than happy to pay the $20 admission fee.

That said, the strategy carries risks. Once the novelty fades and the free tier eventually arrives on mobile as promised, OpenAI will have to work hard to retain these impulse subscribers. If real-world performance fails to match the flashy launch demos, cancellations could follow just as quickly.

For the broader industry, this shatters the idea that mobile is merely an auxiliary interface for generative AI. Phones are where emotional and multimodal interactions happen, and whoever builds the best conversational experience on mobile will capture the consumer's wallet first.