OpenAI's quarterly revenue crossed $10bn for the first time in the three months ending in September, rising nearly 50% from the prior quarter, according to a person familiar with the matter cited by The Straits Times (straitstimes.com). The figures were shared with investors on a call on 10 November. TechCrunch (techcrunch.com), citing Bloomberg reporting based on anonymous sources, put the prior quarter at $6.7bn, implying an almost 50% jump.
The same disclosure carried a less comfortable number. According to documents viewed by The Wall Street Journal and reported by Euronews (euronews.com), OpenAI told investors its operating losses widened to $10.9bn from $10.1bn in the previous quarter. Its adjusted operating margin improved to -109% from -150%. Revenue is climbing fast; the cost of producing it is climbing too.
What the numbers actually show
The revenue figure is a genuine acceleration. TechCrunch notes that growth between the first and second quarters had been about 18%, so the step up to roughly 50% marks a sharp change in trajectory. The Straits Times reports revenue was nearly threefold the same period a year earlier. TechCrunch also reports OpenAI has reached an annual revenue run rate of $50bn. The margin figure needs care. An adjusted operating margin of -109% means operating costs still exceed revenue by more than the revenue itself, even after the improvement from -150%. The narrowing is real progress on unit economics; it is not profitability. Euronews frames the disclosure against investor scrutiny of AI company finances, with Anthropic's impending listing serving as a public test of the sector's spending boom.Why growth resumed
The Straits Times attributes the rebound to competitive new models, aggressive price cuts and a greater focus on AI agents that help customers automate work. TechCrunch describes a similar mix: tighter pricing, discounts, improving models and more enterprise-focused products. Both point to the same commercial mechanism. OpenAI is buying back momentum with lower prices and more useful products, which lifts volume and revenue while adding pressure to the cost base behind inference. That matters for buyers. Discounting and agent tooling are the levers OpenAI is using to win accounts, and those terms are negotiable in a way they were not when the company had fewer credible rivals.The Anthropic comparison
The competitive picture is more nuanced than a single quarter suggests. TechCrunch reports Anthropic recorded third-quarter revenue of $17bn and an annual run rate of $81bn as of the end of October, up $10bn in a month. The Straits Times reports Anthropic's third-quarter revenue at about $17bn but notes its growth rate slowed to around 47% from more than 140% in the previous quarter. The two companies do not measure the same thing. TechCrunch notes Anthropic counts sales made by its cloud partners, while OpenAI reports only its share of some partner sales. Any direct comparison of headline revenue should carry that caveat.Funding and the public market test
OpenAI is completing a funding round at a $1.25 trillion valuation after delaying plans to go public until 2027, according to The Straits Times, which reports an initial $11bn tranche has closed. Anthropic is targeting a $1.4 trillion valuation for a listing later in November. Euronews reports Anthropic aims to sell 825 million shares at between $70 and $78 each under the ticker "ANT", with seven tech companies including Google, Nvidia, AMD and Microsoft agreeing to buy $26bn of non-voting shares at the IPO price. Art Hogan, chief market strategist at B. Riley Wealth Management, told AFP, as reported by Euronews, that the listing is "just another chapter" in investors' worries over the cost of the AI buildout and how it gets financed. He added that "there are three massive companies that investors have wanted to be part of for the last three years and that's SpaceX for sure, Anthropic is next, and perhaps next year OpenAI".Commercial impact
For enterprise buyers, the practical read is that OpenAI's willingness to discount is likely to persist while it defends share against Anthropic and cheaper open-source models. TechCrunch notes competition from Meta, Amazon, Microsoft and Google, plus capable low-cost models from open-source labs, particularly in China. Procurement teams negotiating multi-year contracts have more leverage than a year ago, but should expect pricing to be tied to usage and agent deployment rather than flat seats. For investors and suppliers, the widening operating loss alongside rising revenue confirms that scale alone has not resolved the cost structure behind inference and compute. Capacity commitments remain the binding constraint on how aggressively OpenAI can price.Risks and unknowns
The revenue figures are not public filings. The Straits Times relies on a person speaking anonymously; TechCrunch relies on Bloomberg's anonymous sources; Euronews relies on documents viewed by The Wall Street Journal. None of the numbers has been confirmed by OpenAI directly, and the company's accounting for partner sales differs from Anthropic's, limiting like-for-like comparison. The durability of the growth also depends on whether discount-led demand holds once pricing stabilises.What happens next
Anthropic's listing will provide the first public market valuation of a leading model developer and a reference point for OpenAI's own 2027 plans. OpenAI's funding round and its ability to narrow losses further will be watched as evidence of whether the current growth rate is sustainable at prevailing prices.Sources and References
- TechCrunch — techcrunch.com
- The Straits Times — straitstimes.com
- Euronews — euronews.com
