Anthropic’s IPO Math: $4.6B Revenue, $8B Loss, and a Cheaper Claude Haiku 5.5

yesterday / 22:57 2 sources neutral

Key takeaways:

  • Anthropic's $8B loss underscores AI compute costs, a risk for AI tokens like FET, TAO.
  • Cost-efficient models like Haiku 5.5 may shift value to inference layers, pressuring GPU-reliant crypto projects.
  • Watch RNDR as Anthropic's compute obligations could validate decentralized compute demand, but hype risk persists.

Anthropic's IPO filing has laid bare the brutal economics of frontier AI development. The Claude developer generated approximately $4.6 billion in revenue during 2025, but spent about $7.3 billion on compute and infrastructure alone. Total operating expenses reached roughly $12.65 billion, leading to an operating loss of about $8.06 billion.

The company also disclosed approximately $518 billion in future cloud, computing, and infrastructure obligations, underscoring the immense capital required as Claude usage expands. Anthropic has separately entered arrangements under which Broadcom could finance up to $42 billion of Anthropic chip leasing.

Despite those losses, growth remains explosive. Anthropic's annualized revenue run rate reportedly reached about $65 billion by July 2026, after crossing roughly $47 billion in May. Yet run-rate revenue is not profit, and the company still needs enormous computing capacity to serve demand.

On October 7, Anthropic released Claude Haiku 5.5, completing a product lineup that includes Sonnet 5.5 and Opus 5.5. The new model is about 75% cheaper than Haiku 4.5 on average. For requests up to 100,000 tokens, pricing is $0.10 per million input tokens and $0.50 per million output tokens. Anthropic also reduced cache-read costs for Sonnet 5.5. Early users cited concrete gains: Asana said latency improved by more than 30%, while Box reported an 11-point improvement over version 4.5.

The launch comes as enterprise AI spending accelerates. Gartner estimates spending on AI models and platforms will rise from $39 billion in 2025 to $64 billion in 2026, a jump of more than 60%. A BCG study found nearly half of businesses are already seeing value returns from AI, increasing pressure for cost-effectiveness. A World Economic Forum analysis of 44 supplier invoices found a mid-level AI model handled all invoices, while a more powerful model solved only 91%, suggesting task suitability matters more than raw benchmark scores.

Anthropic has gained ground on OpenAI among US business customers, according to Ramp data covering more than 70,000 businesses. But loyalty remains thin: 52% of businesses using Anthropic or OpenAI are customers of both, and switching between providers reached a record 8% in September.

The key question ahead of a possible IPO after the November 2026 US midterm elections is whether lower-cost models can drive enough usage to create sustainable profits while Anthropic continues spending heavily on AI infrastructure. Reuters reported the offering could value the company at more than $2 trillion.

Previously on the topic:
Oct 2, 2026, 2:43 a.m.
Broadcom to Lend Anthropic Up to $42 Billion for AI Infrastructure
Disclaimer

The content on this website is provided for information purposes only and does not constitute investment advice, an offer, or professional consultation. Crypto assets are high-risk and volatile — you may lose all funds. Some materials may include summaries and links to third-party sources; we are not responsible for their content or accuracy. Any decisions you make are at your own risk. Coinalertnews recommends independently verifying information and consulting with a professional before making any financial decisions based on this content.