Steve Eisman Says Anthropic’s Growth Fell From 147% to 45% Ahead of Its IPO
Hedge fund legend Steve Eisman spotted a crack in Anthropic's growth story just days before its Nasdaq debut, and his efficiency metric may be the single number that determines whether ANT becomes a mania or a milestone.
Steve Eisman, host of The Real Eisman Playbook podcast, said on October 29, 2026, that Anthropic’s sequential revenue growth fell from 147% in the second quarter to 45% in the third. He blamed “the end of token maxing,” meaning heavy usage that boosts usage-based revenue until it fades.
Anthropic’s S-1, filed October 26, 2026, for a Nasdaq listing under “ANT,” leaves third-quarter revenue blank. AMD (NASDAQ:AMD | AMD Price Prediction), an Anthropic supplier, fell 2.08% on October 29 to $566.42, though nothing ties the move to his remarks.
What Eisman’s Numbers Rest On
Revenue climbed through 2025, from $376 million to $2.040 billion via $787 million and $1.382 billion, then jumped from $4.731 billion in March to $11.555 billion in June, about 144% per the filing. Eisman’s 147% matches press rounding to $11.6 billion.
At an October 14 investor day, Anthropic reportedly told prospective investors third-quarter revenue reached $16.8 billion, up more than 45%, or about 45.4% against the filing’s June figure. The account came from an unnamed attendee, and Anthropic declined to comment.
His efficiency gauge is revenue over compute and infrastructure expense: “The ratio in the second quarter was 2.08 times, and the ratio in the third quarter was 1.7.” The first matches the filing’s $5.558 billion of June compute expense, but the second is his estimate. He also shared a reported 50% October price cut he said he “can’t confirm.”
What a Slowdown Means for AMD
AMD disclosed on August 4, 2026 that Anthropic will deploy up to 2 gigawatts of MI450 Series GPUs in Helios racks starting in the first half of 2027, with no dollar value or unit count disclosed. OpenAI and Meta Platforms (NASDAQ:META) each committed to 6 gigawatts.
For a $924.67 billion company at 213 times earnings, the practical risk would be Anthropic stretching out its build-out.
Where the Filing Pushes Back
The filing says June-quarter revenue grew 1,368% year over year while compute expense grew 300%, after the ratio climbed from about 0.69x in December 2025 and 1.40x in March.
Net revenue retention, a measure of existing customers’ spending growth, was about 900% as of September 30, 2026. More than 11,900 customers spent $100,000 or more over twelve months, more than 1,900 topped $1 million and about 250 topped $10 million. About 79% of third-quarter revenue came from outside the top five customers, and revenue per employee topped $6.6 million.
June also brought a first operating profit of $974 million but a $66.3 billion net loss, mostly a $59.6 billion non-cash fair-value change on convertible notes, derivatives, and SAFEs tied to the rising valuation.
Claude Opus 5.5 cut cost per task by about 40% versus Claude Opus 5 through lower prices and efficiency, which can increase usage but also reduce revenue per task.
What Investors Need to Know Before ANT Starts Trading
Anthropic is reportedly seeking up to $100 billion, and some investors reportedly value it near $1.5 trillion, down from $2 trillion but above May’s $965 billion post-money valuation. Trading could reportedly begin before Thanksgiving on November 26, and no price range has been filed.
Eisman calls parts of the prospectus “extremely negative,” yet on a broad market decline said, “I’m not willing to make that bet just yet.” That restraint gives his narrower critique weight.
A reported $1.5 trillion valuation likely leaves little buffer for ANT shares at or soon after the IPO if growth keeps slowing, and riding a mania like this is fine as long as you plan the exit, which is the whole point of our free bubble survivor’s handbook.
Investors researching Anthropic’s demand can also study AMD’s public results, since Anthropic is one of its three anchor customers.
Check the amended S-1 for third-quarter compute expense. A ratio holding near 2.08 times would likely weaken Eisman’s case, while one near 1.7 would likely support it.
Contact [email protected] for any questions or corrections.
Omor Ibne Ehsan is a writer at 24/7 Wall St. He is a self-taught investor with a focus on growth, cyclical, and dividend equities that have strong fundamentals, value, and long-term potential. He also has an interest in high-risk, high-reward investments such as penny stocks.
