OpenAI revenue run rate spurs tech stock drop

OpenAI recently shared investor materials showing that its revenue was approaching $50 billion on an annualized basis at the end of September. This internal figure is about $20 billion lower than an earlier estimate of roughly $70 billion that had circulated among investors the previous month. The $50 billion run rate is a key reference point for assessing demand for OpenAI’s large language model services and the scale of its commercial operations.

The new disclosure places OpenAI behind rival Anthropic on this specific metric. Anthropic has reported an annualized revenue run rate of about $65 billion, implying a higher current revenue pace than OpenAI when using the latest stated figures. The contrast between the two numbers has drawn investor attention amid intense competition among leading AI model developers.

The roughly $20 billion discrepancy between OpenAI’s updated figure and the higher $70 billion estimate stems from differences in how annualized revenue is calculated. In attempting a like-for-like comparison with Anthropic, some investors used a method that mirrored Anthropic’s approach. Anthropic’s run-rate metric includes sales recorded through major cloud partners such as AWS and Google Cloud.

By contrast, the $50 billion value that OpenAI presented to investors did not include comparable cloud-partner revenue. Using a narrower definition of what counts toward the run rate produced a lower number than the figure derived under Anthropic’s broader methodology. The episode underlines how sensitive headline revenue run-rate metrics can be to underlying accounting choices and inclusions.

Following reports of OpenAI’s lower annualized revenue run rate, technology and AI-linked stocks came under pressure. Intraday trading saw the Nasdaq index fall in a range of roughly 1% to 1.6%. The weakness was concentrated in names tied closely to AI infrastructure and semiconductor demand.

Shares of Nvidia (NVDA), a key supplier of AI chips, declined by around 2.9% to 3% during the session. Some infrastructure and enterprise software providers experienced even sharper moves, with Oracle (ORCL) shares dropping by about 5%. Several other chipmakers also traded lower, reflecting investor concern that AI revenue expectations may have been set too high based on earlier, less precise run-rate assumptions.

The adjustment in OpenAI’s run-rate profile has sharpened scrutiny of valuation and growth assumptions across the broader AI ecosystem. Investors are reassessing how to interpret headline revenue metrics when comparing AI model providers that rely on differing mixes of direct and cloud-partner distribution. The contrast between OpenAI’s roughly $50 billion run rate and Anthropic’s reported $65 billion highlights these methodological fault lines.

For AI hardware and infrastructure companies whose fortunes are tied to sustained growth in model inference and training workloads, the updated OpenAI figure has introduced a note of caution into the market. While the numbers still imply very large revenue bases, the reaction shows that even modest recalibrations in perceived demand can move major technology and chip stocks in the short term.