OpenAI's Annualized Revenue Doubles to $40 Billion Ahead of IPO

OpenAI's Run Rate Doubles to $40 Billion as IPO Preparations Accelerate

OpenAI's annualized revenue has climbed past $40 billion, roughly double the run rate the company reported at the end of 2025, according to Bloomberg reporting that cited people familiar with the matter. President and co-founder Greg Brockman told employees Thursday that the annual revenue run rate grew more than 20% month-over-month in July. Chief Financial Officer Sarah Friar had previously put the figure at more than $20 billion for the close of 2025.

The jump is being driven by three areas: rising take-up of AI coding tools, growth in subscription revenue, and the early expansion of an advertising stream. Demand has also increased for OpenAI's agent products, including the coding-focused Codex and the broader workplace tool ChatGPT Work. Friar and board chair Bret Taylor told employees that GPT-5.6, ChatGPT Work and Codex were the main sources of momentum.

The revenue acceleration is unfolding alongside a build-out for a public listing. OpenAI submitted a confidential IPO prospectus to the U.S. Securities and Exchange Commission in June, and the company has just changed its enterprise sales leadership: Dali Rajic, formerly president and chief operating officer of Wiz, is the new chief revenue officer, replacing Denise Dresser. The move follows several senior departures in recent months.

The commercial momentum does not yet translate into profit. OpenAI posted a net loss of $38.5 billion in 2025 on $13.07 billion in revenue, according to audited financial documents. The gap between fast-growing revenue and sizeable losses sits at the center of the company's pre-IPO story.

The IPO Math Behind OpenAI's Revenue Surge and $38.5 Billion Loss

Why Codex, ChatGPT Work and Advertising Are Driving Growth

OpenAI's reported acceleration is not evenly spread across its business. The company's own leadership has named GPT-5.6, ChatGPT Work and Codex as the products behind the July momentum. That matters because these are enterprise and developer-oriented offerings rather than only consumer ChatGPT subscriptions. Advertising is still described as an early revenue stream, so the current surge likely rests more on software and subscriptions than on ads.

Anthropic's $47 Billion Run Rate and the Pricing Fight

OpenAI is competing directly with Anthropic for business customers. Anthropic disclosed a run-rate of more than $47 billion in May, but Bloomberg cautioned that the two companies may calculate that figure differently. OpenAI has also lowered prices on select models as it pursues budget-sensitive buyers in a crowded market that includes Chinese competitors. Board chair Bret Taylor acknowledged that OpenAI had been playing catch-up in coding, which suggests the pricing moves are partly defensive.

The $38.5 Billion Loss Complicates the IPO Narrative

A $40 billion annualized revenue run rate is impressive, but OpenAI's audited 2025 results showed a net loss of $38.5 billion on $13.07 billion of revenue. The discrepancy reflects heavy spending on computing capacity, model development and expansion. For public-market investors, the key question is not just how fast revenue grows, but how unit economics improve as Codex, ChatGPT Work and advertising scale.

Leadership Change Points to Enterprise Sales Focus

Bringing in Dali Rajic from Wiz to lead revenue suggests OpenAI is prioritizing experienced enterprise and security sales leadership before the IPO. The replacement of Denise Dresser and other recent senior departures add execution risk, however. A new revenue organization must convert technical momentum into enterprise contracts at a moment when competition and price pressure are both rising.

What the $40 Billion Run Rate Means for Investors, Enterprise Buyers and Rivals

For enterprise buyers:

• OpenAI has already lowered prices on select models as it competes with Anthropic and Chinese rivals. Use the current coding and agent pricing pressure to negotiate multi-year terms, but verify whether the discounted model versions meet your workload requirements.
• Codex and ChatGPT Work are the products OpenAI itself credits for July's more than 20% month-over-month run-rate growth. Customers evaluating AI agents should test them against actual enterprise workflows rather than general benchmark claims.

For investors and analysts:

• Treat the $40 billion run rate as a revenue-scale signal, not a profit signal. The audited 2025 net loss of $38.5 billion on $13.07 billion of revenue means margin trajectory will be the central S-1 question.
• Do not compare OpenAI's run rate directly with Anthropic's $47 billion disclosure without checking methodology; Bloomberg itself cautioned the calculations may differ.
• Track Dali Rajic's first enterprise sales targets and whether senior departures continue. Leadership churn in a pre-IPO revenue organization is a concrete execution indicator.

For rival AI providers:

• OpenAI's admission that it was playing catch-up in coding, combined with selective price cuts, shows where the competitive opening remains. Competing on developer workflow depth and total cost of ownership matters more than matching headline model pricing.