Anthropic passed OpenAI on revenue — and enterprise coding agents are why

The AI revenue race flipped this year, and the reason matters more than the scoreboard. Anthropic reached a roughly $47B annualized revenue run-rate — up from ~$30B earlier in 2026 — overtaking OpenAI's reported ~$24–25B. The engine behind the swing isn't chatbots. It's enterprises putting AI coding agents to work.

What actually happened

Anthropic's growth has been near-vertical: from an $87M run-rate in early 2024 to tens of billions two years later, per VentureBeat. The standout product is Claude Code, its agentic coding tool, which hit $1B in annualized revenue within six months and $2.5B+ by February 2026. Anthropic also confirmed 1,000+ enterprise customers spending $1M+/year. The strategic tell, as CNBC noted, is focus: Anthropic chased the enterprise while OpenAI courted consumers with ChatGPT.

Why it matters for your business

Read past the billion-dollar numbers and there's a signal that lands whether you run an AI lab or a three-store retail operation: the companies with the biggest engineering budgets are spending real money to make software get built faster. Coding agents aren't a demo anymore. They're a line item.

Here's the part nobody selling you a subscription says out loud: a raw coding agent doesn't run your business. It writes code fast, confidently, and sometimes wrongly. It has no idea how your inventory moves, what your payroll approval flow is, or which webhook can't be allowed to fire twice. The leverage is real — but only in the hands of someone who ships production systems and knows where the human checkpoints belong.

That's the wedge we work from. We use these tools every day — which is why we can build a custom internal tool, Shopify app, or automation in weeks instead of quarters, and hand you something that survives contact with a real Friday afternoon. The speed gets passed to you. The judgment stays in the loop.

Sources: VentureBeat, CNBC.