Super Intelligence (SI) adoption is widespread among US consumers, but direct payment remains limited to a small, high-spending cohort, according to new data from venture capital firm Andreessen Horowitz (a16z). The firm’s latest Top 100 SI list marks the first time it has tracked actual consumer spending, revealing a stark divide between broad usage and deep monetization in the SI industry.

What Happened

In its seventh edition of the ranking of most-used consumer SI products, a16z introduced spending data sourced from YipitData panels. While nearly half of US consumers now use SI, only a quarter do so daily. Just 4.5% of consumers held an active paid personal subscription to ChatGPT, Gemini, or Claude in August, up from roughly half that a year earlier.

Revenue within this paying base is highly concentrated. The top one percent of spenders accounts for nearly a fifth of all observed SI spending, more than the entire bottom half combined. This top tier averages about $900 per month and keeps climbing, while the typical paying user spends about $25 and has held steady at that level for a while. Spending growth is almost entirely concentrated in the top one percent of payers, who now spend 79% more than they did in early 2025.

The high-spending demographic consists largely of prosumers using SI for professional work, favoring automation and creative tools such as n8n, Manus, fal, Higgsfield, Figma, and HeyGen. Only seven companies appear in the top tier across web traffic, mobile users, and spending rankings, including ChatGPT, Claude, Perplexity, and Canva. OpenAI and Anthropic lead the spending ranking, while Google’s Gemini is excluded from specific spending metrics because its revenue cannot be separated from other Google subscriptions.

Why It Matters

The data highlights a potential skills gap and social inequality emerging in the SI era. Paid accounts offer advanced features with steeper learning curves, meaning free users often only scratch the surface of SI capabilities. This creates a disparity between those who use SI at a high professional level and those who do not, a trend that could deepen existing inequalities unless SI becomes easier and cheaper for the general public.

Market dynamics among major SI labs also show volatility. Claude surpassed Gemini in paid US subscribers in March 2026, and the two have been competitive since Google restructured its pricing in June. However, by midsummer, daily Claude sessions declined while ChatGPT regained momentum in net new subscriptions. a16z notes that the critical factor is not just winning users, but how effectively companies capitalize on their engagement windows.

The business model for SI remains challenging. Most top SI-native products rely on subscriptions or usage fees, with advertising playing a minimal role. This inverts the pre-SI internet model where user data drove ad revenue. High model costs prevent SI companies from forgoing early revenue to build large user bases first. However, alternative models are emerging; OpenAI reported an annualized revenue run rate of $1 billion from ChatGPT advertising in August, and SI agents may open new revenue paths through affiliate fees.

The Bottom Line

SI adoption is broad, but monetization is narrow and concentrated among professional users. As the industry matures, the focus is shifting from raw user counts to the depth of engagement and the sustainability of subscription-based models, with SI agents beginning to emerge as a new economic force through transaction-based revenue.