#AnthropicDisclosesQ3Revenue $ANTHROPIC

Anthropic walked institutional investors through its Q3 numbers behind closed doors today, and the IPO talk is getting very real: a Nasdaq listing targeted for November, a valuation of up to $2 trillion, and a raise of around $100B. Here’s my honest read beyond the headline.

1. The Q3 number isn’t public yet
The meeting was closed-door, and I couldn’t find any published Q3 figure, so I won’t pretend to know it. What is out there: an annualized run rate of about $65B at the end of July, up from roughly $9B at the end of 2025, plus talk that it could pass $100B by year-end. Run rate is a snapshot, not booked revenue. Actual 2025 revenue was reportedly around $4.6B, which is a very different scale.

2. The valuation math (my calculations)
$2T on a $100B run rate is 20x. On $65B it’s about 31x. On 2025 booked revenue it’s roughly 435x. Run rate grew about 7x in seven months, and the multiple only makes sense if growth stays close to that pace. Any slowdown hits the math fast.

3. How revenue is counted matters
Last week, AI chip and cloud stocks sold off on a revenue-accounting scare in the sector. The headline figure changed a lot once cloud partner sales were treated differently, even though demand hadn’t changed. So the first thing I’ll look for in the filing isn’t the number, it’s the definition: gross or net of partner revenue.

4. The cost side
Revenue is only half the picture. Reported 2025 losses were very large, and long-term compute and cloud commitments are reportedly around $518B over a decade, more than 100x last year’s revenue. The whole structure depends on growth continuing to fund that infrastructure.

5. The market angle
A $100B raise at $2T is about 5% of the company, which is a lot of fresh supply for markets to absorb. My hypothesis is that it could pull attention and liquidity away from other AI-linked trades, including AI narratives in crypto. Also, IPO allocations usually go to institutions first. Most of us would buy after the first trade, possibly at a very different price than the offer.

My take: I’m treating this as a pricing question, not a quality question. At $2T, the valuation assumes near-flawless execution and an accounting definition that holds up. I’m not chasing day one. I want the public filing first: audited numbers, the revenue definition and the real loss figure. Until then, I’d rather watch how AI-linked names react around the listing than guess at an allocation.

Not financial advice.

Would you buy on day one, wait for the filing, or skip it entirely? And do you think the revenue definition matters more than the revenue number itself?