Anthropic Tops OpenAI in LLM Revenue Share

According to Counterpoint data reported by The Register, Anthropic led global LLM revenue in Q1 2026 with a 31.4 percent share, narrowly ahead of OpenAI at 29 percent. Per Counterpoint as reported by The Register, Anthropic had roughly 134 million monthly active users compared with about 900 million for OpenAI, and a much higher average monthly revenue per active user at $16.20 versus $2.20 for OpenAI. The Register also reports Counterpoint figures of $5 per user for Microsoft, $1.10 for Google, and $0.10 for Meta. The Register quotes Counterpoint describing Anthropic as having "successfully captured the high-end professional market."

Industry evidence shows two monetization approaches in the current LLM market: scale-driven engagement and higher monetization per user. Companies that focus on consumer-scale engagement typically prioritize broad distribution, low friction access, and ad or ecosystem revenue models. By contrast, firms extracting higher per-user revenue tend to sell premium, enterprise, or professional offerings where customers accept per-call or subscription pricing for reliability, compliance, or output quality.

Editorial analysis: The Register highlights a tension practitioners should note: platform-scale growth does not automatically translate into direct monetization per user. The coverage also cites large infrastructure spending by major cloud and platform players, which raises the effective cost base behind scale-focused strategies, while a smaller set of vendors appears to be converting AI into a traditional software revenue stream.

Editorial analysis: Observers should track three indicators to see how this split evolves:

• •changes in published ARPU or revenue-share estimates from firms such as Counterpoint or other analysts;
• •product pricing and feature segmentation that targets enterprise customers versus mass users;
• •quarterly disclosure of infrastructure and R&D spend by major cloud and platform providers, which affects the economics of scale-based models.

For practitioners: these patterns matter for product design, pricing experiments, and choices about latency, reliability, and compliance investments when targeting enterprise versus broad-consumer use cases.