5 Key Numbers From Anthropic's S-1 Filing

Anthropic concluded its S-1 filing, revealing key figures regarding its financials and business strategy. The company plans to list on the Nasdaq under the ticker ANT, aiming for a valuation of up to $2 trillion and potentially raising $100 billion. Here are five crucial numbers from Anthropic's S-1 filing.

Firstly, Anthropic generated $1.85 billion in revenue during the second quarter of 2026, a significant increase from $787 million in the same period the previous year. Within the first half of 2026, Anthropic's revenue reached nearly $16.3 billion, a 12.5-fold increase from the same period in 2025. This surge in revenue is primarily driven by Claude, Anthropic's flagship language model, which is used by consumers and businesses through subscriptions or integration into their products and workflows.

Secondly, Anthropic outlined its substantial investments in data center infrastructure. The company has committed to procure additional Tensor Processing Units (TPUs) from companies like Google between 2028 and 2031, with estimates suggesting these commitments could cost between $125 billion and $150 billion. This spending includes other deals, though some may be canceled on short notice.

Thirdly, the revenue generated by customers outside of Anthropic's top five customers amounted to $9.5 billion during the third quarter of 2026. No single customer accounted for more than 10% of the company's revenue in the nine months ending September 30, 2026, highlighting the diverse customer base.

Fourthly, Anthropic boasts an impressive gross margin of over 70% for the first half of 2026. This high margin indicates robust pricing power, surpassing Nvidia's typical margins of around 70%. However, investors may question whether this margin accurately reflects all costs associated with running these models, such as ongoing training expenses.

Lastly, the S-1 filing is unusually extensive, with over 400 pages dedicated to risk factors. This section raises concerns about AI safety, including the potential for models to act in unintended, misaligned, or autonomous ways, which could harm the company's reputation and financial condition. Despite these risks, the S-1 emphasizes that proper model guardrails and safeguards are crucial for maintaining the company's brand and financial health.