Behind the $852 Billion OpenAI Valuation: Triumph, Pressure, and a Rival Catching Up

When a valuation shifts from aspirational to awkward, a certain silence descends upon a conference room on Sand Hill Road. It’s practically audible. OpenAI is currently experiencing that specific silence after closing the largest funding round in Silicon Valley history at a $852 billion post-money valuation. The money has arrived. The headlines have already been published. The press release is triumphant. However, the question that keeps coming up in the LP emails and group chats is the one that no one really wants to ask aloud. Is this figure truly credible?

In addition to a16z, D. E. Shaw Ventures, MGX, TPG, Fidelity, Sequoia, BlackRock, and a small retail tranche that was the first time individual investors were included through bank channels, OpenAI raised $122 billion in March, led by SoftBank, Amazon, NVIDIA, and longtime backer Microsoft. Retail participation is a crucial detail. It indicates that OpenAI is getting ready to go public, possibly as early as the end of this year. You need at least $1.2 trillion at pricing for an IPO at a valuation that supports $852 billion. For comparison, that is comparable to Eli Lilly today and greater than the peak market capitalization of the majority of oil majors.

To be fair, the operational narrative underlying the figure is truly remarkable. Over 900 million people use ChatGPT every week. In 2025, OpenAI’s annual revenue was $13.1 billion, with a monthly revenue of about $2 billion. Enterprise now makes up more than 40% of revenue, and by year’s end, it is expected to match consumer. The coding agent Codex now serves over 2 million users every week, a five-fold increase in just three months. In most aspects, the flywheel that Sam Altman has been outlining since 2023—consumer adoption driving enterprise deployment driving developer usage driving more compute demand—is genuinely functioning.

However, there is also the ledger’s other column. OpenAI continues to lose money. The projected net loss for 2025 is close to $9 billion. Trillions of dollars will be spent on chips, data centers, and multi-year compute contracts by 2030. To put it politely, the difference between run-rate revenue of about $24 billion and capital commitments exceeding $1 trillion is aggressive. This week, a Reddit commenter compared it to using a $200K salary to max out a million-dollar credit card. Although the analogy is rough, it is not incorrect.

However, competitive pressure is the most intense. From being a cautious second-place research lab, Anthropic has subtly evolved into a significant commercial competitor. Due in large part to Claude’s coding tools, its revenue reportedly increased from $9 billion to $30 billion over the previous year. Secondary-market trades have started to suggest that Anthropic is worth more than $1 trillion; Polymarket is currently pricing about 56 percent odds that Anthropic will formally surpass OpenAI’s valuation before the end of the year. According to a Financial Times story published in mid-April, some of OpenAI’s own investors have started questioning Altman about how $852 billion makes sense when a competitor with a quicker rate of growth might demand more.

Watching this develop gives me the impression that we are at one of those times when investors’ narratives about a company that will define a generation begin to falter, not because the business is failing but rather because the expectations have finally caught up with the facts. OpenAI is not in danger. Its revenue growth is genuine. Adoption of its product is genuine. Microsoft is not abandoning its previous $13 billion investment. However, whether ChatGPT continues to expand won’t be the focus of the upcoming year. They will focus on whether OpenAI can stay ahead of Anthropic without going over budget, reorient toward enterprise without losing the consumer magic, and persuade public markets that $852 billion was a floor rather than a ceiling. Which side of that question will age better is still up for debate.