OpenAI says ChatGPT has hit 20M paid subscribers, up from 15.5M at the end of 2024, suggesting the company is making at least $415M in monthly revenue
Before today's main column about the AI attention war between OpenAI and Google, there's OpenAI news from my colleague Natasha …
Context & Ripple Effects
ChatGPT’s paid base had already moved from more than 11 million subscribers in September 2024 to 15.5 million reported by the end of 2024. The new figure extends that monetization trajectory by another 4.5 million subscribers in a short period.
The growth matters because OpenAI had said consumer subscriptions accounted for roughly three-quarters of revenue, while its prior reported free-to-paid conversion rate was only 5% to 6%. Paid-user expansion is therefore a central measure of whether consumer AI usage can become recurring revenue.
First-order effects
- OpenAI gains a larger recurring subscriber base: 20 million paid ChatGPT accounts imply at least $415 million in monthly revenue under the article’s calculation.
- The increase from 15.5 million at year-end indicates that ChatGPT’s subscription business is still adding customers rather than relying only on its existing base.
Second-order effects
- The result raises the bar for consumer AI rivals, including Google, to show that assistant engagement can translate into paid products or other durable revenue streams.
- A larger paid base gives OpenAI more scope to fund product delivery and customer support, while making the economics of serving subscriber usage more consequential.
Third-order effects
- If sustained, this points to consumer AI shifting from an audience-growth contest toward a subscription-scale contest, where conversion and retention matter alongside usage.
- The durability of that model will depend on whether recurring subscription revenue continues to cover the cost of useful AI tasks as usage expands.
The trend: Consumer AI leaders are increasingly being judged on their ability to convert broad assistant usage into recurring subscription revenue at scale.