Anthropic just had the best quarter in its history — and credit card data proves it.

An analysis of billions of anonymized transactions from roughly 28 million U.S. consumers, conducted by consumer analytics firm Indagari for TechCrunch, shows Claude gaining paid subscribers at record levels throughout Q1 2026. Anthropic confirmed to TechCrunch that paid subscriptions have more than doubled this year.

The surge wasn’t driven by one event. It was a compounding sequence of moves that turned Anthropic from a developer-focused company into a household name.

The Three Catalysts

1. The Super Bowl Gambit (February)

Anthropic aired commercials during the Super Bowl mocking ChatGPT’s decision to show ads to users — and promised Claude would never do the same. The spots were funny, effective, and visibly irritated Sam Altman.

New subscribers flooded in. The majority landed on Claude Pro at $20/month — the lowest paid tier.

2. The Pentagon Feud (January–March)

Starting in late January, reports emerged of a deepening clash between Anthropic and the Department of Defense. Anthropic refused to allow Claude models for lethal autonomous operations or mass surveillance of American citizens. The DoD threatened consequences, eventually designating Anthropic a supply chain risk.

CEO Dario Amodei’s firm public statement on February 26 became a catalyst. Indagari’s data shows new user growth climbing sharply between the initial media reports in late January and Amodei’s statement.

The contrast couldn’t have been starker: OpenAI announced a Pentagon deal; ChatGPT uninstalls surged 295%. Users voted with their wallets.

3. Claude Code, Cowork, and Computer Use (January–March)

Product releases sealed the deal:

  • Claude Code (January) — a developer tool that became one of the most talked-about AI coding environments of the year
  • Claude Cowork (January) — productivity tooling that brought Claude Code’s capabilities to non-developers
  • Computer Use (this week) — Claude can now navigate a computer independently: clicking, scrolling, and executing tasks. Combined with Dispatch (task assignment from phones), this makes Claude a genuine autonomous agent

These features are paid-only, giving users a concrete reason to subscribe beyond ideology.

The Hard Numbers

The Indagari data reveals several key patterns:

  • Record new subscribers between January and February 2026
  • Record returning users in February — people who had previously tried and left Claude came back
  • Growth continuing through early March (the latest available with a two-week data delay)
  • Majority entering at Pro tier ($20/month), suggesting mainstream consumer adoption rather than just power users

This is meaningful because it’s based on actual credit card transactions — not app downloads, not engagement metrics, not company press releases. Real money, real commitment.

Still Behind ChatGPT

For all the momentum, Claude remains significantly behind ChatGPT in absolute market share. OpenAI is still gaining new paid subscribers at a rapid pace and remains the biggest consumer AI platform.

There’s also a sobering counterpoint: Anthropic’s overall market share (across the full AI landscape, not just consumer) has actually declined from 29.1% in March 2025 to 13.3% in March 2026, partly due to competition from cheaper Chinese AI models, according to The Register.

The consumer surge is real, but Anthropic is winning a battle within a war that’s getting more crowded by the month.

What This Means for OpenClaw Users

The Claude consumer boom has direct implications for the OpenClaw ecosystem:

Capacity pressure is real. On March 26, Anthropic adjusted usage limits for free, Pro, and Max subscribers during peak hours (5 AM–11 AM PT), allowing users to hit their 5-hour session caps faster. About 7% of users were hitting limits. If you’re running Claude through OpenClaw with API access, you’re on separate infrastructure — but expect the API pricing conversation to evolve as Anthropic manages demand across consumer and API channels.

Claude Code integration deepens. Claude Code’s popularity means more developers are building tools that work with Claude’s agent capabilities. OpenClaw users running Claude Code as a coding agent benefit from this expanding ecosystem of integrations and community knowledge.

Computer Use changes the agent game. Claude navigating computers independently is exactly the kind of capability OpenClaw was built to orchestrate. As Computer Use matures, expect tighter integration patterns — your OpenClaw agent delegating screen-level tasks to Claude’s Computer Use, supervised by your own safety policies.

Model diversity matters more than ever. With Claude’s consumer demand straining capacity, and market share data showing the competitive landscape fragmenting, OpenClaw’s model-agnostic design becomes increasingly valuable. Run Claude when it’s the best fit, switch to alternatives when it’s congested or when a cheaper model handles the task.

The Bigger Picture

What’s happening with Claude isn’t just about one company doing well. It’s evidence that consumers will pay for AI products that align with their values — and that trust is a competitive moat.

Anthropic bet that refusing the Pentagon would cost them government revenue but win them something harder to quantify: public trust. The credit card data suggests that bet is paying off, at least for now.

For the AI agent ecosystem, this matters because the platforms where users concentrate their spending and attention become the platforms where developers build. More Claude subscribers mean more Claude-compatible tools, more Claude-optimized workflows, and more community knowledge — all of which benefits OpenClaw users who run Claude as their primary model.

The AI platform race isn’t just about benchmarks anymore. It’s about who users trust with their $20 a month.