OpenAI Revenue Run Rate Tops $40 Billion, Doubling From 2025

OpenAI’s revenue run rate is reportedly on track to top $40B in annualized terms, indicating a step-change in scale. The article also notes this would represent a doubling versus 2025 levels, which supports a constructive demand and monetization outlook for the business.

Analysis

The important signal is not the absolute revenue level; it is that AI monetization is still scaling fast enough to justify continued capex from hyperscalers and model-infrastructure vendors. That keeps the immediate winners anchored in the picks-and-shovels layer: NVDA, AVGO, MSFT, ORCL, and the data-center/power ecosystem (DLR, EQIX, VRT, ETN). The second-order read is that if one frontier model vendor can grow this quickly, enterprise procurement teams are likely still prioritizing capability over price, which reduces the odds of a near-term demand air pocket.

The risk is that revenue growth may be compute-expense growth in disguise. If inference costs are rising almost as fast as sales, the market could eventually re-rate the model layer lower while keeping infrastructure multiples elevated; that is a relative-value negative for high-multiple AI software names with vague monetization paths. Over the next 1-3 months, the real catalyst is not this print itself but hyperscaler capex commentary and any evidence that token pricing, usage growth, or gross margins are inflecting.

Contrarian view: the market may be overgeneralizing one company’s scale into a broad AI bull case. The structurally durable trade is not "AI software" broadly; it is power, networking, and GPU supply where demand is more visible and less substitutable. The thesis fails if capex guides flatten, if open-source/on-prem models compress pricing, or if regulatory friction slows deployment over the next 6-18 months.

Key Decisions for Investors

• Prefer a relative-value long NVDA / short IGV basket for the next 1-3 months; the market is likely still underpricing infrastructure pull-through versus app-layer monetization. Falsify if hyperscaler capex guidance rolls over or NVDA order momentum slows.
• Add to MSFT on pullbacks as the cleanest public-market beneficiary of durable AI spend; use a modest size because the upside is real but already partly reflected in valuation. Reassess if Azure growth decelerates or OpenAI-related spend becomes a margin drag.
• Initiate a 6-18 month long basket in DLR, EQIX, VRT, and ETN to express the power/cooling/data-center buildout rather than model-layer enthusiasm. The risk is that a faster-than-expected model commoditization cycle reduces incremental infrastructure intensity.
• Do not chase the long-tail AI software trade here; consider short exposure to the highest-multiple names with weak monetization proof points (for example SNOW/ADBE-type exposure) only if upcoming earnings fail to show tangible AI revenue contribution. Cover if AI attach rates or billings accelerate.
• Set an alert for the next round of hyperscaler capex and OpenAI partner disclosures; if capex growth pauses or token prices fall materially, trim AI infrastructure longs quickly because the market will de-rate the whole stack on margin compression fears.