Only 4.5% of US consumers pay for AI, and the top 1% spends $900

Share
Only 4.5% of US consumers pay for AI, and the top 1% spends $900

Consumer AI has near-half household reach and a paying base the size of a rounding error — and today's numbers on who actually spends put a fine point on it.

Only 4.5% of US consumers pay for AI, and the top 1% of them spends $900 a month. Andreessen Horowitz published the seventh edition of its Top 100 generative AI apps ranking, and for the first time the firm tracked observed spending on US consumer cards alongside traffic and downloads. The picture: nearly half of US adults now use AI, about a quarter use it daily, but just 4.5% had an active paid subscription to ChatGPT, Gemini, or Claude in August — roughly double the 2.1% of a year earlier, still a narrow base for an industry priced like this one. Within that base, money is brutally concentrated: the top 1% of payers accounts for 19.5% of observed spending (more than the entire bottom half combined), averages about $903 a month, and has grown 79% since early 2025, while the median payer sits flat at around $25. Their purchases tell you who they are — automation and build tools like n8n, Manus, and fal, plus creative software like Figma and HeyGen. These are prosumers buying a workflow, not casual chat users, and it suggests consumer AI's real market so far is professionals one subscription away from a business expense.

The ranking itself shows a market settling at the top. ChatGPT leads Gemini by roughly two to one in web visits and Claude by six to one, and it holds about three times either rival's US paid subscribers. Claude overtook Gemini in US paid subscribers in March 2026 and has run neck and neck since Google's June pricing change — though daily Claude sessions slipped in July and August while ChatGPT's net-new subscriptions pulled ahead again. Only seven products appear across the spending, web, and mobile top-50 lists (ChatGPT, Claude, Suno, Perplexity, Photoroom, Canva, and Notion), and the edition drew just 11 new entries, the fewest of any of the seven reports. On the business-model side, 84% of leading AI-native web products charge subscriptions and 64% charge by usage, against just 14% using ads — a16z's read is that high model costs stop labs from giving AI away to build an audience, inverting the ad-funded pre-AI internet. With the panel caveat that card data is a slice rather than total revenue, the sobering takeaway is that AI monetization is real but extremely top-loaded: a small, professional minority is funding a consumer industry everyone else uses for free.

What to watch: whether the 4.5% keeps doubling on schedule or flattens as the easy early adopters are exhausted.

Is the prosumer tier the only durable consumer AI market — or does something cheaper and simpler unlock the other 95%? Tell us in the comments.

Read more

Google's Gemini 4 Carbon reportedly matches Opus 5.5 on coding

Google's Gemini 4 Carbon reportedly matches Opus 5.5 on coding

Google hasn't launched Gemini 4 Argon yet, but internal documents suggest a faster follow-up is already in testing — and early impressions put it level with Anthropic's best coding model. Google is testing a Gemini 4 variant called "Carbon" that employees say performs on par with Anthropic's Opus 5.5 on programming tasks. Business Insider, citing internal documents, screenshots, and chats, reports that Google deployed Carbon on Jetski — its internal coding platform — over the past few days, wi

Open Source Radar — October 10: reviewers, skills, and 3D maps

Open Source Radar — October 10: reviewers, skills, and 3D maps

Today's trending board skews practical: tooling that fixes code, teaches agents a framework, and rebuilds 3D scenes from video. Here's what's worth your stars. alibaba/open-code-review (Go, 45.6k stars) — Alibaba open-sourced the code review tool it runs internally, and it shows: a hybrid of deterministic static-analysis pipelines for the known bug classes (null dereferences, thread-safety, XSS, SQL injection) plus an LLM agent that writes precise, line-level comments. It talks to OpenAI- or A

Deep Dive — Congress has the data center numbers, still no bill

Deep Dive — Congress has the data center numbers, still no bill

A yearlong Senate investigation into seven of the biggest data center developers in the country concluded this week that the public case for the AI buildout does not survive the companies' own paperwork — and it landed at the exact moment Congress needs it, because the one federal bill written to make data centers pay for their own power fell three votes short of advancing in the Senate, despite passing the House 417-3. The report, led by the offices of Senators Elizabeth Warren, Chris Van Holle

WorldArena 2.0 puts world models to the test on real robots

WorldArena 2.0 puts world models to the test on real robots

The world-model field has argued for two years about video quality. This week the first full results landed for a benchmark that asks the harder question: can a robot actually use the prediction? WorldArena 2.0's global challenge has closed its leaderboard, and for the first time world models are being graded on physical robots instead of plausible video. The benchmark — designed by a Tsinghua-led consortium with PKU, CMU, Stanford, Princeton and others — extends its 1.0 video scoring along th