Anthropic’s annualised revenue hit $30 billion in April 2026, surpassing OpenAI’s $25 billion for the first time and marking the fastest revenue ramp in enterprise software history. The company grew from $9 billion ARR at end-2025 — roughly tripling in four months.
What’s driving the surge
Over one thousand businesses are now spending more than $1 million per year on Anthropic’s Claude AI models. Enterprise adoption has accelerated since early 2026, with buyers making multi-year commitments and expanding contracts at rates the company describes as unprecedented. The revenue is coming from coding assistance, customer support automation, and increasingly, autonomous agent deployments across finance, legal, and manufacturing.
Infrastructure to match
To sustain the demand, Anthropic has locked in a 3.5-gigawatt compute deal with Google and Broadcom — securing multiple gigawatts of next-generation Tensor Processing Unit capacity. The agreement represents one of the largest compute procurement deals in history and ensures Anthropic can scale its frontier Claude models without the GPU supply constraints that have bottlenecked competitors.
Claude Mythos: too powerful for public release
Anthropic’s latest model, Claude Mythos Preview, has been restricted to a select group of enterprise partners — Amazon, Apple, Cisco, Google, JPMorgan Chase, and Microsoft — specifically for cybersecurity defence purposes. The company has determined that Mythos is not ready for public launch because of the ways it could be exploited by cybercriminals and state-sponsored threat actors. It’s a notable departure from the “ship fast” mentality that has defined the AI race.
Building its own chips
In a move that signals long-term ambition, Anthropic is in early discussions about designing its own AI chips, according to Reuters. The plans are preliminary — no dedicated team or specific design has been committed — but reflect a broader industry shift as frontier AI labs seek to reduce dependency on external chip suppliers.
Leave a Reply