Anthropic Revenue Surges 14-Fold as Company Posts First Positive Operating Income

Anthropic’s revenue exploded more than 14-fold in the second quarter from a year earlier, reaching preliminary figures above $11.5 billion, according to documents viewed by Bloomberg News. The Claude chatbot maker doubled its first-quarter revenue of $4.73 billion and surged from $787 million in the same period last year, marking a dramatic acceleration as the company prepares for a potential blockbuster initial public offering.

The San Francisco-based AI company also posted positive adjusted operating income in the second quarter, according to the documents. The figures remain preliminary and could still change before the company finalizes its IPO paperwork. An Anthropic spokesperson did not immediately respond to a request for comment.

The jump in revenue comes as Anthropic competes with OpenAI for corporate customers and gains traction among professionals using its software for tasks including coding. Anthropic said in May that its run-rate revenue crossed $47 billion, up from roughly $10 billion in revenue in all of 2025. People familiar with the company’s finances say Anthropic projects revenue of roughly $190 billion to $200 billion for 2028, dwarfing the annual run rate the company shared just months ago.

Enterprise Revenue Powers Margin Turnaround

The company’s achievement of positive operating income while maintaining hypergrowth represents a pivotal moment for the AI industry, which many observers wrote off as a commodity race to zero just a year ago. Roughly 80% of Anthropic’s revenue now comes from API and enterprise usage rather than chatbot subscriptions, reportedly across more than 300,000 business customers. Enterprise API revenue carries structurally higher margins than consumer freemium models because it is metered, scales with customer usage, and carries almost no acquisition cost once a workload is embedded.

Anthropic is not selling seats-it is selling consumption. This shift in revenue mix explains the margin turn better than the raw growth numbers alone. The company has found a monetization path that converts raw model capability directly into metered, high-margin API calls, and it has moved furthest down that path among frontier AI labs.

Claude Code Drives Embedded Enterprise Adoption

The agentic-coding product Claude Code serves as the engine behind Anthropic’s enterprise growth, jumping from roughly $500 million in run-rate revenue in September 2025 to over $2.5 billion by February 2026. The product reportedly sits inside more than a thousand million-dollar-plus enterprise accounts, creating deeply embedded relationships that drive recurring revenue. Coding demand carries a rare property-it converts developer workflows into sustained API usage that becomes mission-critical to customer operations.

This embedded usage creates switching costs that protect revenue streams and improve unit economics over time. As developers integrate Claude Code into their daily workflows, usage naturally expands across teams and projects without additional sales effort. The product demonstrates the cleanest monetization path any AI lab has discovered, translating model performance directly into measurable business value for enterprise customers.

Inference Efficiency Gains Drop Through to Margins

Reported unit compute costs fell from around 71 cents to 56 cents, representing a roughly 20% efficiency gain that drops almost straight through to gross margin. In a business where the marginal cost of revenue is inference, a fifth off the unit cost creates immediate operating leverage. Unlike traditional software companies that spread fixed costs over more users, AI companies also benefit from continuous improvements in compute efficiency as models and infrastructure improve.

These efficiency gains compound over time as the company optimizes its inference stack, negotiates better hardware deals, and develops more efficient model architectures. The combination of growing revenue and falling unit costs creates a powerful margin expansion dynamic that few observers predicted for frontier AI companies.

IPO Preparations Advance Amid Strong Financials

Anthropic’s early meetings with prospective investors ahead of its potential IPO have been high-level and have not included discussions about specific financials or a valuation, sources told CNBC’s David Faber. Anthropic CFO Krishna Rao is leading the meetings, the sources said. The company reportedly plans to go public at a valuation near $2 trillion in late September or early October, which would make it one of the first major private AI companies to tap public markets.

An IPO as soon as this fall would unlock access to billions of dollars in additional capital, providing the immense financial runway required to fund escalating compute infrastructure costs, secure advanced hardware, and build specialized data centers. The public offering would also provide liquidity for early investors and employees while giving the company a currency for future acquisitions and strategic partnerships.

Growth Numbers Challenge AI Bubble Narrative

Anthropic’s explosive growth muddies the narrative that AI represents an overinflated bubble destined to collapse. The company’s revenue more than doubled in a single quarter, even as skeptics warn about unsustainable valuations and unproven business models. Anthropic says it grew revenue by more than 10x in each of the three years leading up to early 2026, demonstrating sustained momentum rather than a temporary spike in demand.

If these numbers hold up, they signal that demand for proprietary AI services continues to surge even amid political headwinds and growing competition from China. Financial services firm Ramp recently measured a slight flattening in demand for Anthropic tokens among its business customers, suggesting some moderation in growth rates ahead. However, the company’s ability to maintain hypergrowth while turning profitable challenges the view that AI labs face inevitable margin compression.

Nvidia Deal Reduction Contrasts with Anthropic Momentum

The strong financial performance at Anthropic contrasts with recent signs of caution from other AI ecosystem players. Nvidia and OpenAI are reportedly closing in on a data center deal, but Nvidia cut its initial guarantee nearly in half under investor pressure. Instead of backing $250 billion as originally planned, Nvidia would initially guarantee just under $120 billion, according to the Wall Street Journal. Investors pushed back over concerns about Nvidia’s risk exposure in the deal.

The guarantee covers the first construction phase, which would deliver about five gigawatts of capacity. OpenAI is separately negotiating a lease for the full 10-gigawatt project being developed by SB Energy, a SoftBank subsidiary. Nvidia is also in talks over separate financing for OpenAI’s chip purchases worth up to $350 billion. The scaled-back deal could give ammunition to critics who have warned about an AI bubble, as Nvidia halving its risk under investor pressure might signal that even the biggest beneficiaries of the AI boom are growing more cautious about future commitments.