A difference in revenue-reporting methodology explains much of the gap, as intensifying competition with Anthropic puts AI companies’ growth, profitability and financial transparency under the spotlight

OpenAI has reportedly informed investors that its annualised revenue run rate for September 2026 was approaching $50 billion, below an earlier indication that the figure was nearing $70 billion. The discrepancy has drawn attention to how leading artificial intelligence companies measure growth as they prepare for potential public listings and face rising expectations from investors.

According to a person familiar with the matter, the difference largely stems from an effort to make OpenAI’s revenue figures directly comparable with those of rival Anthropic. The latest figure was first reported by the Financial Times. OpenAI did not respond to a request for comment, according to the reported account.

The gap does not necessarily indicate that OpenAI’s business has contracted. Instead, it highlights how accounting practices, distribution partnerships and the definition of revenue can significantly influence comparisons between fast-growing AI companies.

Why OpenAI’s Revenue Estimate Changed

OpenAI had previously indicated at a separate investor event that its revenue run rate for September was approaching $70 billion. The more recent figure of nearly $50 billion reflects a different basis of comparison, with the treatment of revenue generated through cloud partners emerging as a key distinction.

OpenAI does not include certain revenue from sales through cloud partners, including Amazon Web Services and Google Cloud, in the same way Anthropic accounts for its business. This difference makes headline revenue figures difficult to compare without understanding the underlying methodology.

The distinction is important because AI companies increasingly rely on major cloud platforms to distribute their products, provide computing capacity and serve enterprise customers. Depending on the commercial arrangement, the amount recognised as revenue by an AI developer may differ from the total value of services sold through a partner.

Consequently, investors need to distinguish between revenue recognised by a company, the value of transactions processed through distribution channels and the broader commercial activity associated with its products.

OpenAI’s Revenue Has Grown Rapidly

Despite the revised comparison, OpenAI’s reported revenue trajectory points to substantial expansion in demand for artificial intelligence services.

The company began 2026 with an annualised revenue run rate of approximately $20 billion, compared with around $6 billion in 2024. This growth reflects the expanding use of generative AI tools across consumer applications, software development, business productivity and enterprise services.

OpenAI’s business model spans paid subscriptions, enterprise products, API access for developers and commercial arrangements that help businesses integrate AI capabilities into their operations. The ability to convert widespread adoption into recurring revenue will remain a major factor in determining its long-term financial performance.

However, a rapidly rising revenue figure alone does not establish whether the business is profitable. Training and operating advanced AI models require substantial computing infrastructure, specialised chips, energy, research investment and technical talent. Investors will therefore be interested not only in sales growth but also in operating expenses, cash requirements and the economics of serving each customer.

Anthropic’s Revenue Growth Intensifies Competition

Competition between OpenAI and Anthropic has become an increasingly important part of the AI industry's financial narrative.

OpenAI reported quarterly revenue of approximately $6.7 billion in the second quarter, while Anthropic reported $11.5 billion, according to the figures cited in the report. This marked the first time Anthropic's quarterly revenue exceeded OpenAI's on the reported comparison.

Anthropic's annualised revenue reportedly crossed $65 billion in July, with sources previously indicating that it was on track to reach $100 billion by the end of 2026. These figures underline the speed at which demand for advanced AI products is expanding, particularly in enterprise applications and AI-assisted software development.

Nevertheless, direct comparisons require caution. Different revenue-recognition policies, cloud distribution agreements, product mixes and billing arrangements can affect reported results. Revenue leadership over a single quarter also does not automatically establish which company has the stronger long-term competitive position.

For OpenAI, the challenge is to maintain customer growth, strengthen enterprise adoption and expand monetisation while competing against specialised AI providers and technology companies with significant infrastructure and distribution capabilities.

Cloud Partnerships Complicate Financial Comparisons

Cloud platforms play a central role in the economics of the AI industry. They provide computing infrastructure and can act as distribution channels for AI products, particularly for enterprise customers that already purchase cloud services.

Anthropic reportedly pays cloud partners approximately 16% of every dollar earned through those channels. Sales through such partners accounted for about half of its revenue last year, according to a Reuters analysis cited in the report.

Arrangements of this kind can create differences between the gross value of sales and the revenue ultimately recognised by a company. They may also affect margins because a portion of the commercial proceeds goes to infrastructure or distribution partners.

For investors evaluating AI businesses, the relevant questions include how much revenue is retained after partner payments, how dependent a company is on external infrastructure and whether its customer relationships generate sustainable recurring income.

The growing importance of these agreements also means that headline revenue rankings should be assessed alongside disclosures about accounting policies and the commercial terms governing partnerships.

Potential IPOs Could Bring Greater Financial Transparency

OpenAI and Anthropic are preparing for potential public listings, according to the report. An IPO process would bring greater scrutiny of their financial performance and could provide investors with more detailed information about revenue composition, costs, cash flows and business risks.

Public-market investors typically assess more than a company's growth rate. They also examine the quality and predictability of revenue, customer concentration, margins, capital expenditure, contractual commitments and the path towards sustainable profitability.

For AI companies, infrastructure spending will be particularly significant. Demand for advanced computing capacity can rise sharply as customer usage increases, but the cost of supplying that capacity may also place pressure on margins. Businesses will need to demonstrate that revenue growth can eventually support their operating and investment requirements.

Greater disclosure could also help investors understand how much growth comes from subscriptions, enterprise contracts, API usage and cloud-based distribution arrangements. These details would be important when determining appropriate valuations for companies operating in a rapidly evolving market.

Why Annualised Revenue Run Rate Can Be Misleading

Annualised revenue run rate is a commonly used metric among high-growth technology companies. It often involves taking revenue generated over a short period, such as one month, and multiplying it by 12 to estimate an annual figure.

While the calculation offers a quick snapshot of business momentum, it is not the same as actual annual revenue. It assumes that the measured month's performance can be sustained throughout the following year.

That assumption can be unreliable when a business is expanding rapidly, experiencing seasonal demand, signing large contracts or changing its pricing structure. One unusually strong or weak month can materially alter the resulting estimate.

The metric also does not reveal whether revenue is profitable, whether customers will renew their subscriptions or how much capital is needed to support future growth. Differences in accounting and distribution arrangements can further complicate comparisons between companies.

Investors should therefore treat annualised run rates as indicators of momentum rather than definitive measures of financial strength. Actual reported revenue, cash flow, margins and customer-retention data provide a more comprehensive basis for assessing performance.

Market Outlook

The gap between OpenAI's earlier indication of nearly $70 billion in annualised revenue and the latest figure approaching $50 billion highlights a broader challenge for investors assessing the artificial intelligence industry: rapid growth figures can be difficult to interpret without consistent reporting standards.

OpenAI's expansion from a reported $6 billion annualised run rate in 2024 to approximately $20 billion at the beginning of 2026 points to strong commercial momentum. However, its comparison with Anthropic demonstrates why investors must look beyond headline numbers and examine how revenue is recognised, how cloud partnerships are structured and how much of the business can translate into sustainable earnings.

Potential IPOs could become a major test of investor confidence in the AI sector. More detailed financial disclosures may help markets distinguish between companies with strong customer demand and those capable of converting that demand into durable profitability.

For investors, the key indicators to monitor will be actual quarterly revenue, operating margins, infrastructure expenditure, cash flow, enterprise customer growth and the consistency of financial reporting. Until comparable disclosures become available, annualised revenue estimates should be interpreted cautiously rather than treated as a definitive ranking of the industry's leading AI companies.