Anthropic’s annualized revenue surges to $65B

Anthropic’s annualized revenue run rate surged to over $65B at end-July, up from $47B in May and $9B at end of last year, signaling an accelerating AI growth trajectory. Investors expect the company to finish 2026 at roughly $100B–$120B, while it targets a potential $2T+ public valuation ahead of a likely IPO ahead of OpenAI (possibly as soon as this fall), which would rank among the largest market debuts on record. The article also notes OpenAI’s revenue doubling to $40B from $20B, but emphasizes Anthropic’s growth metrics capturing investor attention more strongly; Anthropic’s last valuation was $965B after a $65B funding round in late May.

Analysis

The market implication is less about Anthropic itself and more about the durability of AI capex: if private revenue is still inflecting this fast, hyperscalers and GPU vendors get another quarter or two of cover to keep spending aggressively. The cleanest beneficiaries are the infrastructure layer — NVDA, AVGO, TSM, and the cloud platforms — because they monetize usage and capacity before anyone can prove end-demand quality at the model layer.

The second-order loser is the public software cohort that has been trading on an AI re-rating without matching the revenue slope. A headline-grabbing IPO would likely pull capital toward a narrower set of perceived winners and make it harder for slower-growth names in IGV/ARKK-style baskets to justify premium multiples; that effect can show up immediately in flows and over 1-3 months in relative performance, even if fundamentals do not change. If the debut prices at a very high revenue multiple, it also raises the bar for every listed AI name to defend its own monetization path, not just its product story.

Contrarian view: the consensus may be overconfident that top-line growth at a frontier model company converts into durable economics. Private-market run-rate numbers are noisy, often mix usage growth with customer concentration, and can be front-loaded by a small set of large contracts; if growth normalizes after the IPO window, the air pocket could hit AI beta hard. The key falsifier is any sign that hyperscaler capex guides down, model-training spend stalls, or the IPO gets delayed because public investors demand a materially lower multiple than the private round implied.

Key Decisions for Investors

• Long NVDA vs short IGV for the next 4-8 weeks: express the view that AI spending remains infrastructure-led while software multiples are vulnerable to a capital-reallocation trade; target 10-15% relative outperformance if the IPO window stays hot.
• Add to SMH or a basket of NVDA/AVGO/TSM on pullbacks ahead of the filing window: favorable 3-6 month risk/reward if private AI revenue momentum keeps hyperscaler bookings elevated; falsify if capex commentary rolls over.
• Avoid chasing high-multiple AI software names into the IPO hype cycle; use any post-debut rally in CRWD, SNOW, MDB, or NET to trim exposure if the market starts preferring 'picks-and-shovels' over application-layer growth.
• Set an alert on the Anthropic IPO filing and pricing range: if the valuation implies a very rich 2026 revenue multiple, expect a short-term supply event in AI beta and consider tactical hedges via QQQ puts or NVDA call spreads.
• Watch cloud spend indicators over the next 1-3 earnings cycles from AMZN, GOOGL, and MSFT; if they confirm acceleration, the trade remains long infrastructure, but if spend growth decelerates, unwind AI beta quickly.