Anthropic IPO Prospectus Leaked: $42B Loss, Over Half A Trillion In Cloud Commitments — Skeptics Lash Out

Anthropic’s financials lay bare the enormous capital demands behind the AI race.
Anthropic co-founder and CEO Dario Amodei speaks at the INBOUND 2025 event on September 04, 2025 in San Francisco, California. (Photo by Chance Yeh/Getty Images for HubSpot)
Anthropic co-founder and CEO Dario Amodei speaks at the INBOUND 2025 event on September 04, 2025 in San Francisco, California. (Photo by Chance Yeh/Getty Images for HubSpot)
Profile Image
Yuvraj Malik·Stocktwits
Published Sep 28, 2026   |   10:32 PM EDT
Share
·
Add us onAdd us on Google
Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...Loading...
  • Anthropic’s losses soared even as revenue grew 12-fold, highlighting the extreme cost of frontier AI development.
  • The company reportedly has $518 billion in future cloud, computing and infrastructure commitments.
  • The disclosures come as Anthropic rolls out new Claude models and prepares for a potential IPO after the U.S. midterms.

Advertisement|Remove ads.

Anthropic PBC is burning cash at a staggering pace even as it faces tens of billions of dollars in cloud and infrastructure commitments over the coming year, Reuters reported Monday, citing its review of the AI company’s yet-to-be-released initial public offering prospectus.

Anthropic lost $42 billion on revenue of $4.6 billion last year, according to the report, with losses surging 425% from 2024 even as revenue grew 12-fold over the same period. About $7.33 billion was on compute and infrastructure last year, a threefold surge from 2024

Read Next
Loading...
Loading...

The company lost more than $8 billion on an operating basis, excluding write-downs of various liabilities mostly tied to previous fundraising. The IPO prospectus reportedly lists commitments totaling $518 billion on cloud, computing, and infrastructure obligations in the coming years.

Advertisement|Remove ads.

The numbers offer the clearest look yet at the economics and scale of one of the world’s biggest and most influential AI companies, underscoring the enormous costs involved in developing and running frontier AI systems.

Anthropic’s IPO is now likely to come after the U.S. midterms in November, according to Reuters.

Analysts Question Anthropic’s Massive Spending Plans

Ross Hendricks, an equity analyst and author of the Ross Report, questioned whether Anthropic’s future infrastructure commitments were sustainable given their sheer scale.

Advertisement|Remove ads.

“Just a casual mention of raising half a trillion in a single year, really? Does anyone believe this? Let's say they pull $100B in equity from the IPO... and then what, $400B from the debt markets?” he said in an X post.

Hendricks separately questioned Anthropic’s reported $42 billion net loss, arguing that the figure comes as the company faces increasing competition from lower-cost open-source AI models.

EZ Primary Research CEO and AI critic Ed Zitron took a similarly skeptical view of Anthropic’s economics, focusing on its spending relative to revenue. “Bahahahaha Anthropic is a total dog of a company. Spent $12.6bn to make $4.6bn in revenue in 2025, $7.33bn of which was compute costs.  Operating loss of $8bn. Losses getting worse,” he said in an X post.

Advertisement|Remove ads.

Zitron has previously argued that AI companies' economics depend heavily on continued access to capital and spending by major cloud providers.  

Anthropic Stuns With New Claude AI Models

The financial details emerged just as Anthropic rolled out its new Claude Opus 5.5 model last week and Claude Sonnet 5.5 on Monday. 

Opus 5.5 is Anthropic’s flagship model for complex coding, agentic workflows, research and knowledge work, with a 1-million-token context window. Anthropic positions it for tasks requiring deeper reasoning and judgment, putting it in direct competition with OpenAI’s GPT-6 Astra and other frontier models. Claude Sonnet 5.5 is Anthropic’s latest mid-tier frontier model.

Advertisement|Remove ads.

In a new analysis, benchmarking outlet Artificial Analysis put Opus 5.5 at 58 on its Intelligence Index at maximum effort, the highest score it has measured. The model leads six of the index’s 10 benchmarks. Opus 5.5 ranks narrowly ahead of GPT-6 Astra.

Artificial Analysis gave Sonnet 5.5 a score of 56, just two points behind Opus 5.5 and making it the second-highest-scoring model in its maximum-effort tests. 

Anthropic’s private market valuation is $1.32 trillion, according to Nasdaq Private Market, with retail investors on Stocktwits placing a ‘neutral’ stance on the company as of late Monday.

Advertisement|Remove ads.

“Anyone else feel like Anthropic is reaching the point where dragging out the IPO is doing more harm than good?” a trader asked on Stocktwits.  

“It’s one thing to be careful and deliberate when pushing a $2T valuation. And the money is surely out there for them to get it done.  But if they keep dragging things out, seems like the chances of it flopping somewhat when trading starts will go up due to handwringing and worry.”

For updates and corrections, email newsroom[at]stocktwits[dot]com.

Advertisement|Remove ads.

Comments
Share your thoughts...

Comments posted here will also appear on symbol pages.

Follow on Google News
Read about our editorial guidelines and ethics policy

Advertisement|Remove ads.