Claude Code Passes a $2.5 Billion Run Rate Less Than Nine Months After Launch
Anthropic says its terminal-based coding agent more than doubled its revenue pace in six weeks, sharpening the fight with Cursor, Codex and Copilot.
Key takeaways
- Anthropic said Claude Code's run-rate revenue exceeded $2.5 billion, more than double its level at the start of 2026.
- Weekly active users of the tool also doubled since January 1, and Anthropic raised $30 billion at a $380 billion valuation.
- Rivals report on different metrics, so the AI coding race is hard to compare directly.
Anthropic announced that Claude Code, the coding agent it opened to the general public in May 2025, now has a run-rate revenue above $2.5 billion. The company said that figure has more than doubled since the beginning of 2026, a pace that makes the tool one of the fastest-growing software products on record.
What Anthropic said
The disclosure came alongside Anthropic's $30 billion Series G round, which valued the company at $380 billion after the money. In the announcement, Anthropic said Claude Code's run-rate revenue grew to over $2.5 billion and that its weekly active users had doubled since January 1, about six weeks earlier. Coverage of the announcement reported that business customers accounted for more than half of that revenue.
As always, run rate is not revenue booked. It takes a recent period and multiplies it up to a year, so it flatters fast-growing products and can fall if usage drops. Some outlets also treat the number as an estimate rather than an audited figure, which is why the claim is best read as Anthropic's own characterization of momentum.
A race measured in different units
Comparing Claude Code with its rivals is awkward because each company reports something different. Cursor said its annualized revenue topped $1 billion in November 2025. OpenAI has framed Codex's growth by users: reports put weekly users at roughly 1 million in early March and above 4 million by May 2026. Microsoft has said GitHub Copilot passed 20 million users in mid-2025, with about 4.7 million of them paying by January 2026, according to secondary reports.
The only safe conclusion is directional. Agentic coding, where a tool plans work, edits many files and runs tests with limited supervision, moved from novelty to a main line of business for the model labs within a year. Anthropic sells the model and the product, which gives it unusual leverage: its tool can improve in lockstep with its model releases.
Why it matters
For engineering leaders, the signal is that budgets are moving from seat-based editor plugins toward usage-based agents that do whole tasks. That changes procurement: costs scale with how much work is delegated, not with headcount. It also raises questions about review, since more code is written and merged by tools than any team can read line by line.
Anthropic said Claude Code's run-rate revenue "has grown to over $2.5 billion."
The funding backdrop
The Series G is the context that makes the revenue claim land. A $30 billion raise at a $380 billion post-money valuation values Anthropic as one of the most expensive private companies in the world, and a product line growing this quickly is the kind of evidence investors cite. Claude Code reached the general public only in May 2025, so a run rate above $2.5 billion in February 2026 implies the product went from launch to a multi-billion-dollar pace in roughly nine months.
Growth that fast brings a practical question about durability. Coding agents are sticky once a team wires them into repositories, review workflows and continuous integration, but they are also easy to compare. Developers can run the same task through two tools and judge the results, which keeps pricing pressure on every vendor. Anthropic's advantage is that it controls the model underneath, so improvements in reasoning and long-task reliability arrive in the product without waiting on a third party. Its exposure is the reverse: if a rival model leaps ahead, switching costs for individual developers are low.
The disclosure also tells us what Anthropic wants the market to believe. By singling out Claude Code, the company is positioning itself as an enterprise software business with a flagship product, not only a chatbot maker. Business customers accounting for more than half of the tool's revenue supports that reading.
What to watch
- Whether Anthropic publishes updated Claude Code figures at its next funding or product milestone.
- Margin pressure: heavy agent usage burns model tokens, which can be expensive to serve.
- How Cursor, Codex and Copilot respond with pricing or bundling.
- Enterprise controls, such as audit trails and permission limits, that large buyers demand before widening access.
Sources
- Anthropic: Claude Code run-rate revenue — Simon Willison's feed
- Anthropic hits $380B valuation: $30B funding and Claude Code's $2.5B revenue — NxCode
- OpenAI Codex growth and enterprise AI agents — Fortune
- OpenAI Codex 4M weekly developers growth data — Digital Applied
