Anthropic's Valuation Faces Pressure: Revenue Growth Slows, Some Investors Shift to $1.5 Trillion

marsbitPublished on 2026-10-16Last updated on 2026-10-16

Abstract

Anthropic faces valuation pressure for its potential IPO as its revenue growth slows. In Q3, revenue reached $16.8 billion, a 45% sequential increase, which was a significant drop from the over 140% growth seen in Q2. This slowdown has led some potential investors to suggest lowering the IPO valuation expectations, with discussions now around figures possibly under $1.5 trillion, down from earlier informal bank estimates of approximately $2 trillion. The deceleration reflects increased competitive pressures from OpenAI, Google, Meta, and open-source models. While Anthropic has achieved adjusted profitability for two consecutive quarters and positive cash flow in Q3, concerns are growing. Investors are scrutinizing whether future growth can justify high valuations, especially given Wall Street's expectations for a high annualized revenue run rate. Furthermore, Anthropic's massive planned infrastructure investment—reportedly over $500 billion in chips and infrastructure over the next decade—adds to the focus on whether revenue growth can translate into sustainable cash flow and profit. In a climate of rising financing costs and broader pressure on AI-related assets, the market is applying stricter scrutiny to the balance between growth, investment returns, and capital expenditure for AI companies.

Anthropic's revenue continued to grow rapidly in the third quarter, but the pace of growth slowed significantly, prompting some potential IPO investors to reassess the valuation expectations for this AI company. Against a backdrop of pressure on AI deals and rising financing costs, market doubts are deepening over whether high growth can sustain high valuations.

According to a report by The Information, Anthropic disclosed at an investor conference in San Francisco on Wednesday that its third-quarter revenue reached $16.8 billion, representing a quarter-on-quarter increase of approximately 45%, which is lower than the over 140% growth recorded in the second quarter.

Some potential investors believe underwriters should lower their IPO valuation expectations, suggesting a potential valuation below $1.5 trillion, compared to the roughly $2 trillion level that some investment banks had previously discussed privately.

Competition in the industry is intensifying. OpenAI is strengthening its enterprise AI products and competing with lower prices; new models from companies like Google and Meta, as well as open-source models, are also putting pressure on Anthropic.

Anthropic's growth slowdown adds further uncertainty to AI deals already facing pressure from inflation and rising interest rates. The market is not only focused on the revenue scale of AI companies but is also beginning to scrutinize whether growth expectations can be met and whether massive investments in chips and infrastructure can yield corresponding returns.

Revenue Growth Slows Significantly, Testing Growth Expectations

Anthropic's third-quarter revenue increased to $16.8 billion from $11.6 billion in the second quarter, continuing its significant growth from the first quarter's $4.7 billion. However, the drivers behind the company's previous rapid revenue climb are changing.

Second-quarter revenue grew by over 140% quarter-on-quarter, primarily driven by a surge in demand for Claude Code; the third-quarter sequential growth rate has now fallen to around 45%. Although revenue is still expanding rapidly, this deceleration in growth could affect investors' assessment of the company's future growth trajectory.

There are also changes in the revenue structure. Anthropic estimates that revenue related to model distillation accounts for approximately 5% to 7% of total revenue, and the company expects this portion of revenue to decline in the current quarter, continuing to drag on overall revenue performance. Model distillation refers to other companies using the outputs of existing AI models to train their own models.

Anthropic has achieved adjusted profitability for two consecutive quarters, and its cash flow turned positive in the third quarter. The company disclosed an adjusted operating profit margin of 11% for the year-to-date, but this metric excludes significant costs such as stock-based compensation.

Some Investors Reassess IPO Valuation

The slowdown in revenue growth is putting pressure on Anthropic's IPO valuation expectations.

Some potential investors were disappointed with the latest financial data, believing that underwriters need to lower valuation expectations, with some investors discussing a potential valuation level below $1.5 trillion, lower than the approximately $2 trillion previously suggested privately by some investment banks.

This change reflects a re-evaluation of valuation by investors and does not mean that Anthropic has officially lowered its IPO pricing. For potential investors, the key question is whether the company can sustain sufficiently fast revenue growth to support the previous high valuation expectations.

Jefferies analyst Jeffrey Favuzza noted that Wall Street may expect Anthropic's annualized revenue run rate to reach at least $80 billion, with hopes of exceeding $105 billion by year-end. Market expectations for growth are already very high, so even if revenue continues to grow, failure to meet those expectations could trigger valuation pressure.

Massive Infrastructure Investment Increases Return Pressure

Beyond revenue growth and profit improvement, Anthropic also needs to address massive infrastructure investments. Media previously reported that Anthropic plans to invest over $500 billion in chips and infrastructure over the next decade.

Investments on such a scale make the question of whether revenue growth can be consistently translated into cash flow and profit an important part of the valuation discussion. In an environment of rising financing costs, investors are also paying more attention to whether AI companies can scale their businesses at a reasonable cost, rather than just pursuing revenue size.

Anthropic's financial disclosures have captured market attention at a time when AI-related assets are already facing broader pressure. Concerns exist about the financial health of AI cloud service provider CoreWeave, and South Korean memory chip stocks have experienced significant declines. Meanwhile, the market is also digesting risks associated with hotter-than-expected inflation data and the possibility of interest rates remaining high for longer.

These factors do not necessarily mean that AI demand is weakening across the board, but they are changing how the market evaluates AI companies. For companies requiring continuous large-scale capital investment, the relationship between financing costs, revenue growth, and return on investment is being scrutinized more strictly.

Related Questions

QWhat were Anthropic's Q3 2024 revenue and its growth rate compared to Q2?

AAnthropic's Q3 2024 revenue was $16.8 billion, representing a quarter-over-quarter growth of approximately 45%. This was a significant slowdown compared to the over 140% growth experienced in Q2.

QHow has Anthropic's revenue growth slowdown impacted its potential IPO valuation according to the article?

AThe slowdown in revenue growth has led some potential IPO investors to reassess Anthropic's valuation. They believe underwriters should lower expectations, with discussions now centering on a potential valuation below $1.5 trillion, down from previous private discussions of around $2 trillion.

QWhat specific factor was cited as a key driver of Anthropic's explosive revenue growth in Q2 2024, and why did it change?

AThe explosive revenue growth in Q2 (over 140%) was primarily driven by a surge in demand for Claude Code. The article notes that the forces driving the company's rapid revenue climb are changing, though it does not specify the exact reason for Claude Code demand normalizing in Q3.

QBesides slowing growth, what other major financial challenge does Anthropic face regarding its future plans?

AAnthropic faces the challenge of massive infrastructure investments. The company reportedly plans to invest over $500 billion in chips and infrastructure over the next decade. This raises questions about whether future revenue growth can translate into sufficient cash flow and profit to justify such enormous expenditures, especially in a higher cost-of-capital environment.

QWhat broader market conditions are adding pressure to AI company valuations like Anthropic's?

ABroader market conditions adding pressure include: concerns over other AI-related assets (like cloud service provider CoreWeave and Korean memory chip stocks), hotter-than-expected inflation data, and the risk that interest rates will remain higher for longer. These factors are causing investors to scrutinize the relationship between financing costs, revenue growth, and investment returns more strictly for capital-intensive AI companies.

Related Reads

From Harshly Criticizing Stablecoins to Building an L1: Airwallex's Jack Zhang Isn't "Converting" — the Battle for the Settlement Layer Has Begun

In June 2025, Airwallex CEO Jack Zhang publicly questioned stablecoins, arguing that fiat conversion costs outweighed their cross-border payment benefits and that crypto had failed to solve real-world problems for 15 years. Yet just 17 months later, in November 2026, Zhang emerged as co-founder of Alta, a blockchain project that raised $80 million (a $47.5 million Series A led by Index Ventures plus a $32.5 million seed round). Rather than a change of heart, the move reflects commercial necessity: a battle over the stablecoin settlement layer is underway, and payment giants that stay out risk becoming mere interfaces atop someone else's chain. Alta is not a general-purpose public chain but a compliant Layer 1 built for banks and licensed payment institutions. Co-founder Loong Wang, previously of Ren Protocol, says existing chains like Ethereum and Solana cannot simultaneously meet institutional demands for privacy, compliance, integration with local financial systems, and liquidity. Alta uses recursive zero-knowledge proofs to enable on-chain compliance checks without exposing identities or transaction amounts, and builds in programmable compliance modules, native identity authorization, and anti-airdrop protections. Co-founder Catherine Porter previously led Meta's Libra (Diem) partnerships. Airwallex itself is Alta's key advantage: with $287 billion in annualized volume, $1.3 billion in revenue, 200 countries, and 90 currencies, it will channel enterprise clients onto the new chain. Zhang's apparent reversal is consistent: he criticized stablecoins as cheap remittance tools, but now bets on tokenized fiat settlement infrastructure. Competitive pressure comes from Stripe, which acquired Bridge for $1.1 billion, bought Privy, and incubated the Tempo chain (raising $500 million at a $50 billion valuation, with OpenAI, Shopify, Deutsche Bank, and Visa as partners). Circle's Arc, Fireblocks, and Canton Network are also competing. With stablecoin market cap exceeding $320 billion and 2025 on-chain volume surpassing $33 trillion, the real contest is over who controls minting, clearing, and fiat redemption. Success will depend less on technology than on three factors: genuine adoption by traditional financial institutions, compliance adaptation across jurisdictions, and efficient on- and off-ramps to real banking systems. The paradigm has shifted: instead of disrupting banks, fintech giants now aim to build infrastructure that lets traditional finance voluntarily move on-chain. Pragmatism, not ideology, now dominates the race.

marsbit1h ago

From Harshly Criticizing Stablecoins to Building an L1: Airwallex's Jack Zhang Isn't "Converting" — the Battle for the Settlement Layer Has Begun

marsbit1h ago

"Everyone is too conservative with AI!" Head of Claude Code: My approach is more ambitious

Felix Rieseberg, who leads Claude Code's desktop app, argues that most people use AI too conservatively and should set more ambitious goals. His central trick: set unreasonably aggressive targets and insist they're achievable—for example, instead of cutting website load time from 10 seconds to 5, aim for 0.2 seconds, and when the model protests, keep telling it the problem is solvable and to try again. He organizes ambition into four categories: speed (doing the same work faster), height (raising the abstraction level to focus on desired outcomes rather than individual tasks), depth (working like top experts, using their real tools, data, and methods), and breadth (drawing on ideas from other fields). For height, he suggests asking what you ultimately want to happen and what the goal behind the goal is—e.g., instead of fixing single crashes, require the app to never crash. For a quarterly business review, move from filling in a template to understanding what changed and why, identifying the most urgent decisions, and redesigning the review process itself. For depth, he advises insisting on the tools top professionals actually use (databases and statistical models rather than Markdown tables), setting "unreasonable" targets, challenging your own assumptions against raw data, hunting for new knowledge, and using external validation rather than letting the model grade itself. A useful prompt asks the model what top experts would do that you haven't requested, what raw materials and tools they'd use, and what besides the model and you could prove the result correct. For breadth, he recommends replacing nouns with verbs (not "make a PPT" but "in 30 minutes, convince busy people of an explanation and drive a decision under uncertainty"), assembling cross-disciplinary teams, and using multi-agent setups—while avoiding mere role-play in favor of adopting other fields' actual methods. He closes with self-efficacy research: people given clear, challenging goals outperform those told to do their best, and believing difficult things are possible—as with the four-minute mile—unlocks achievement. In an era of rapid capability gains, updating our beliefs about what's possible matters more than ever.

marsbit1h ago

"Everyone is too conservative with AI!" Head of Claude Code: My approach is more ambitious

marsbit1h ago

The Traditional Santa Rally: Will It Happen in US Stocks This Year?

US stocks have suffered a sharp selloff, but technical and positioning conditions are repairing, with Goldman Sachs, Morgan Stanley and JPMorgan all signaling a potential year-end rally. Goldman strategist Cosimo Codacci-Pisanelli argues that aggressive deleveraging (the most intense in a decade), valuation multiples back near 35-year averages (MSCI world forward P/E at 15.6x), upward EPS revisions, and seasonal factors create conditions for a year-end short squeeze. Morgan Stanley data shows hedge funds are now underweight AI for the first time since 2023, with the longest deleveraging streak in over 15 years; JPMorgan calls poor sentiment a bullish signal. However, the report flags a paradox: when all three banks call the same trade, the contrarian "year-end squeeze" becomes a crowded consensus trade. Worse, inflation pressure is building faster than during the pandemic: global manufacturing import price inflation is running at about 24% annualized, and the share of import categories with annualized gains above 5% has jumped to 25-40% across major economies. Firms are absorbing costs via margin compression, but upstream pressures typically pass through to consumers in 6-12 months, meaning the bill arrives in Q1. January is historically the month with the most price resets, and Goldman has added a third Fed hike to its forecast. A reflexive loop means a rally that loosens financial conditions removes the rationale for central bank caution, effectively borrowing from future returns. Goldman recommends long front-end inflation as a portfolio hedge, and notes the best rule for 2026 is to short every consensus trade.

marsbit1h ago

The Traditional Santa Rally: Will It Happen in US Stocks This Year?

marsbit1h ago

Claude Deciphers Gödel's Cryptic Manuscripts: 80-Year-Old Notes Shake the Mathematics World

In 1978, logician Kurt Gödel left behind thousands of pages of unpublished notes written in Gabelsberger shorthand, a nearly extinct German stenography system understood by only a handful of people. In 2017, Finnish logician Jan von Plato, upon reaching mandatory retirement age, set out to decode 16 workbooks comprising 1,009 pages that Gödel wrote after settling in Princeton in 1940. After years of manually transcribing the shorthand into modern English, his team could barely digest one page per week—until computer scientists suggested using Claude. Fed the transcripts, von Plato's own books, and modern literature, Claude produced a 650-page interpretation within days, later expanding to 1,084 pages. The AI identified 663 theorem-like results, around 440 of which were confirmed as new theorems through formal verification—suggesting Gödel proved nearly one theorem every two days in the early 1940s. Claude also revealed that Gödel came within a hair of proving the independence of the Axiom of Choice two decades before Paul Cohen's 1963 forcing method earned him a Fields Medal; Gödel had constructed an equivalent framework but made an error in a key definition of "finite support," a result later verified in Coq and praised by Gödel expert Akihiro Kanamori. The notes also contain results rediscovered only in the 1970s–80s, 14 unexplored research topics, and physics work including early use of the Lüders rule (proposed in 1951) and serious study of time-travel via his rotating "Gödel universe." Gödel died in 1978 of starvation caused by paranoia.

marsbit1h ago

Claude Deciphers Gödel's Cryptic Manuscripts: 80-Year-Old Notes Shake the Mathematics World

marsbit1h ago

Trading

Spot
活动图片