Skip to main content
TechnologyApr 7, 2026· 2 min read

Anthropic's Revenues Tripled in Four Months, Surpassing Its Direct Competitor

It is undeniable that the fiercest battle currently taking place in the technological landscape is that of artificial intelligence. Within this competition, Anthropic has announced a run-rate of $30 billion, which is the projection of annual revenues based on current monthly results. This figure marks a vertical growth compared to the approximately $9 billion recorded at the end of 2025, with the company having nearly tripled its revenue in just four months, capitalizing on the massive adoption of the Claude family models by the enterprise sector.

The numbers position Anthropic ahead of OpenAI in the financial growth rate metric. The company led by Sam Altman recently reported monthly revenues in the range of $2 billion, achieving an annual run-rate of $24 billion. Although the two companies may adopt slightly different calculation criteria, the trend indicates that Anthropic is eroding market share at a faster rate than its main competitor, largely thanks to a solid customer base: there are over a thousand companies currently spending more than a million dollars a year to integrate Claude APIs into their workflows.

Financial Overtaking: Anthropic Accelerates to $30 Billion

To sustain what is defined as exponential demand, Anthropic has finalized an agreement with Google and Broadcom, which provides access to next-generation computing capacity based on TPU (Tensor Processing Unit) in the multi-gigawatt range. The infrastructure is set to become operational in 2027 and will be crucial for training frontier models and their subsequent global deployment.

The decision to diversify computing hardware is a crucial technical step. While Anthropic continues to use NVIDIA GPUs and AWS Trainium hardware, the significant commitment to Google’s TPUs suggests a strategy aimed at optimizing inference and training costs on custom architectures. Krishna Rao, CFO of Anthropic, described the operation as the most significant computing commitment in the company’s history, necessary to keep Claude at the top of LLM performance.

Despite the new partnership with Google for silicon supply, Amazon remains the primary cloud partner and the main ally for model training. However, Anthropic maintains a position of distributive neutrality, which is a unique feature in the sector. Claude is indeed the only 'frontier' class model available simultaneously on the three major cloud platforms: AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry.

This agnostic availability allows Anthropic to engage business clients regardless of the infrastructure already present in the company. While OpenAI is increasingly focusing on solutions optimized for dedicated coding software agents and a vertical integration with the Microsoft ecosystem, Anthropic seems to be betting on pure scalability and integration flexibility, a wager that, judging by the $30 billion run-rate, is yielding dividends exceeding initial analyst expectations.