Anthropic's $1.4 trillion IPO faces a $901 billion question: Can AI revenues justify its compute bill?
Anthropic's proposed $1.4 trillion stock market debut comes amid soaring revenues and improving profitability, but its massive $901 billion compute commitments are raising questions about long-term sustainability.

- Nov 15, 2026,
- Updated Nov 15, 2026 5:15 AM IST
Anthropic's upcoming initial public offering (IPO) is emerging as a major test of investor confidence in artificial intelligence, with the Claude developer seeking a valuation of approximately $1.4 trillion despite concerns over mounting infrastructure commitments.
According to an AP report, Anthropic has set an IPO price range of $70–78 per share, with trading expected to begin Thursday under the ticker ANT. At the upper end, the offering could raise $64.4 billion.
The proposed valuation is below the $2 trillion figure discussed earlier this year, reflecting growing scrutiny of AI spending.
$901 billion compute commitments raise concerns
A November 13 report by Ashika Institutional Equities identifies Anthropic's approximately $901 billion in compute and infrastructure commitments as a major financial risk.
Of this amount, approximately $452 billion falls due before the end of 2030, while $451.2 billion represents leases signed but not yet commenced.
The report cautions that profitability will depend increasingly on revenue generated from contracted computing capacity.
According to Ashika, Anthropic generated $2.08 in revenue for every dollar spent on compute during the second quarter of 2026. However, the ratio declined to $1.72 in the third quarter as infrastructure expenditure increased.
The brokerage estimates that operating break-even requires approximately $1.20–$1.25 in revenue per dollar of committed compute, under its assumptions.
MUST READ: OpenAI’s revenue hits $10 billion, but its losses are still staggering at $10.9 billion
Revenue jumps 1,206%, profitability improves
Despite these commitments, Anthropic's financial performance has strengthened considerably.
Ashika reported that revenue surged 1,206% year-on-year to $33.24 billion during January–September 2026.
Third-quarter revenue reached approximately $17 billion, translating into an annualised revenue run rate of $68 billion.
Adjusted operating income stood at $4.1 billion during the nine-month period, while free cash flow turned positive at $5.6 billion.
However, Ashika expects compute efficiency to weaken temporarily in early 2027 as additional infrastructure becomes operational ahead of demand.
ALSO READ: Anthropic’s record-breaking $90 billion IPO is already oversubscribed several times over
Amodei bets on enterprise AI
Meanwhile, Anthropic CEO Dario Amodei has emphasised AI safety and enterprise adoption as central to the company's growth strategy.
"We're very paranoid about AI safety," Amodei told JPMorgan Chase CEO Jamie Dimon during an investor event, according to AP.
"We deliberately chose enterprise as a focus because we understood that it was synergistic with our focus on safety," he said.
Amodei also expects AI applications to expand beyond coding into scientific research and other industries.
"AI isn't just one application; it isn't just one market," he said.
For investors, the central question remains whether expanding enterprise demand can generate sufficient revenue to support Anthropic's long-term infrastructure commitments without eroding profitability.
DO READ: Anthropic IPO: Claude-creator targets $1.4 trillion valuation, eyes $90.4 billion fundraise
Anthropic's upcoming initial public offering (IPO) is emerging as a major test of investor confidence in artificial intelligence, with the Claude developer seeking a valuation of approximately $1.4 trillion despite concerns over mounting infrastructure commitments.
According to an AP report, Anthropic has set an IPO price range of $70–78 per share, with trading expected to begin Thursday under the ticker ANT. At the upper end, the offering could raise $64.4 billion.
The proposed valuation is below the $2 trillion figure discussed earlier this year, reflecting growing scrutiny of AI spending.
$901 billion compute commitments raise concerns
A November 13 report by Ashika Institutional Equities identifies Anthropic's approximately $901 billion in compute and infrastructure commitments as a major financial risk.
Of this amount, approximately $452 billion falls due before the end of 2030, while $451.2 billion represents leases signed but not yet commenced.
The report cautions that profitability will depend increasingly on revenue generated from contracted computing capacity.
According to Ashika, Anthropic generated $2.08 in revenue for every dollar spent on compute during the second quarter of 2026. However, the ratio declined to $1.72 in the third quarter as infrastructure expenditure increased.
The brokerage estimates that operating break-even requires approximately $1.20–$1.25 in revenue per dollar of committed compute, under its assumptions.
MUST READ: OpenAI’s revenue hits $10 billion, but its losses are still staggering at $10.9 billion
Revenue jumps 1,206%, profitability improves
Despite these commitments, Anthropic's financial performance has strengthened considerably.
Ashika reported that revenue surged 1,206% year-on-year to $33.24 billion during January–September 2026.
Third-quarter revenue reached approximately $17 billion, translating into an annualised revenue run rate of $68 billion.
Adjusted operating income stood at $4.1 billion during the nine-month period, while free cash flow turned positive at $5.6 billion.
However, Ashika expects compute efficiency to weaken temporarily in early 2027 as additional infrastructure becomes operational ahead of demand.
ALSO READ: Anthropic’s record-breaking $90 billion IPO is already oversubscribed several times over
Amodei bets on enterprise AI
Meanwhile, Anthropic CEO Dario Amodei has emphasised AI safety and enterprise adoption as central to the company's growth strategy.
"We're very paranoid about AI safety," Amodei told JPMorgan Chase CEO Jamie Dimon during an investor event, according to AP.
"We deliberately chose enterprise as a focus because we understood that it was synergistic with our focus on safety," he said.
Amodei also expects AI applications to expand beyond coding into scientific research and other industries.
"AI isn't just one application; it isn't just one market," he said.
For investors, the central question remains whether expanding enterprise demand can generate sufficient revenue to support Anthropic's long-term infrastructure commitments without eroding profitability.
DO READ: Anthropic IPO: Claude-creator targets $1.4 trillion valuation, eyes $90.4 billion fundraise