Anthropic revenue run rate soars to $30 bn as Claude Code drives 80× growth

Anthropic revenue run rate has vaulted to ↑ $30 bn, an escalation that dwarfs the three‑decade climb of legacy software giants. CEO Dario Amodei, a former OpenAI VP, disclosed the figure at the Code with Claude conference, noting that the firm had prepared for a ten‑fold surge but instead witnessed an 80× jump in Q1 usage.

Anthropic revenue run rate fuels enterprise AI boom

The surge is anchored by Claude Code, the agentic coding assistant launched in mid‑2025. Within six months it crossed ↑ $1 bn in annualized revenue and now tops $2.5 bn run‑rate, with weekly active users doubling since January.

“Software engineers are the fastest adopters of new tech,” Amodei told the audience.

Claude Code reads entire codebases, devises action plans, executes them with real development tools, and iterates autonomously, while developers retain final commit authority. The tool now writes the bulk of Anthropic’s internal code, creating a feedback loop that rivals cannot replicate.

Enterprise demand has exploded: over 1,000 customers now spend >$1 m annually on Claude services, a cohort that includes Uber and Netflix. The company’s client base has doubled since February, echoing a broader AI‑driven shift noted by Reuters.

Hyper‑growth has strained compute capacity. To bridge the gap, Anthropic struck a deal with Elon Musk’s SpaceX, securing access to the Colossus 1 data centre’s >300 MW of GPU power – more than 220,000 Nvidia units spanning H100, H200 and upcoming GB200 chips. The partnership follows earlier agreements with Amazon, Google and Microsoft, but the SpaceX deal delivers immediate relief.

Analysts at Bloomberg note that Anthropic’s valuation may soon breach $900 bn, potentially eclipsing OpenAI. A prospective IPO is rumored for late 2026, with Goldman Sachs and JPMorgan already in preliminary talks.

Despite a Pentagon blacklist labeling the firm a supply‑chain risk – a move that could cost billions – the company’s momentum remains unchecked, reminiscent of the post‑pandemic tech acceleration.

Amodei warned that the next seven months will test whether infrastructure, organization and vision can keep pace with demand, as the “country of geniuses” inside data centres grows ever more crowded.

Words by Dr. Aris Thorne (Artificial Intelligence Researcher).