a16z Data: 45% of US Consumers Use AI, Only 4.5% Pay—Top 1% Spend $900/Month

Andreessen Horowitz has published its latest Top 100 AI list, and for the first time the firm tracked actual US consumer spending on AI products rather than relying on download or usage metrics alone. The findings show a stark divide: roughly 45 percent of US consumers now use AI tools, but just 4.5 percent pay for a subscription. The data offers the clearest picture yet of how money actually flows through the consumer AI market.

The spending that does happen is heavily concentrated at the top. The highest-spending 1 percent of paying users shell out about $900 per month, according to a16z's analysis. That group is not buying casual chatbots or novelty image generators; their money goes mostly to professional tools for software development and workflow automation. In other words, the people extracting economic value from AI are the ones willing to pay enterprise-level prices.

That pattern mirrors what has played out in other software categories, but the ratio here is more extreme. Freemium models have driven adoption far faster than they have driven revenue, and the gap between the two is wider in AI than in most consumer software markets. Companies have leaned on free tiers to build habits and gather data, betting that a fraction of users will eventually convert. The a16z numbers suggest that bet is paying off for a narrow slice of the market, not the whole.

The concentration also raises questions about which AI businesses can survive on consumer revenue alone. If only 4.5 percent of users pay, and most of that revenue comes from a tiny professional cohort, then consumer AI startups face a choice: chase the high-spending developers and automation buyers, or find entirely different monetization paths such as advertising, commerce, or bundling into existing subscriptions. The $900 monthly spenders are valuable, but they are also a limited pool.

The a16z Top 100 list itself has become a reference point for tracking which AI apps gain traction, and adding spending data changes what the ranking can tell observers. Usage charts can be inflated by curiosity and churn; payment data is harder to fake. By publishing both, the firm gives founders and investors a way to separate products people try from products people actually need.

What remains unclear is whether the 4.5 percent conversion rate will rise as AI tools become more capable, or whether it will stay flat while spending concentrates further at the top. The answer will shape how consumer AI companies price their products, where they spend on acquisition, and whether the next wave of AI apps is built for millions of casual users or a smaller number of professionals willing to pay serious money.