"43 Minutes Condensed into 10 Key Takeaways: The Essence of Anthropic's Trillion-Dollar IPO Roadshow" On November 10th, Anthropic released a 43-minute IPO roadshow video, showcasing for the first time its astonishing revenue growth, enterprise customer base, and future profitability goals. From a 45-fold increase in revenue over six quarters to a customer net revenue retention rate of up to 900%, and a long-term computing power commitment of nearly $900 billion, what exactly is the business model of this AI giant, valued at approximately $1.4 trillion? We have compiled the 10 most noteworthy key points from the roadshow and the latest prospectus. 1. Quarterly Revenue Soars 45-Fold In the first quarter of 2025, Anthropic's revenue was only $376 million; by the third quarter of 2026, it had reached nearly $17 billion, representing a more than 45-fold increase over six quarters and a year-on-year growth of approximately 12 times. 1. In the first three quarters of this year, the company achieved a gross profit margin of 85%, an adjusted operating profit margin of 12%, and positive free cash flow. 2. Approximately 6% of revenue comes from users who violated usage policies. Anthropic disclosed in its prospectus that approximately 5%–7% of its revenue for the first three quarters of 2026 came from customers subsequently found to have violated usage policies or business terms. The company acknowledged that future revenue growth may be affected as bans become more stringent. 3. Net revenue retention rate as high as 900%. Anthropic disclosed that its enterprise customer net revenue retention rate is approximately 900%. This means that enterprises are not only starting to use Claude, but are also continuously expanding their usage and budgets. 4. More than half of Fortune 500 companies are paying for it. As of the third quarter, Anthropic had more than 11,900 enterprise customers with annual spending of $100,000 or more, nearly eight times the number at the beginning of the year. Over 50% of Fortune 500 companies and 30% of the Global 2000 pay Anthropic more than $100,000 annually. 5. AI Reduces Two-Month Work to 5 Minutes Pharmaceutical giant Bristol-Myers Squibb deployed Claude to 30,000 employees, reducing some data queries that previously took one to two months to complete to just four or five minutes. Insurance company AIG accelerated some of its quote processes by 55%, and Accenture reduced some cybersecurity scanning tasks from several days to less than an hour. 6. AI Has Begun to Help Anthropic Develop AI Anthropic disclosed that it launched 250 models, products, and features in the first nine months of 2026. Compared to 2024, the average daily code contribution per engineer increased 45 times; from January to September this year, the scale of employee use of Claude increased more than 250 times. 7. What used to take dozens of engineers months, AI completed in three days. Anthropic adapted to AMD GPUs for the first time this year. Similar work in the past would have required dozens of engineers and months. With Mythos's involvement, deployment was completed in just three days, with almost all related kernel code written or migrated by Claude. 8. Revenue increased 45 times, while computing power costs only increased 10 times. Over the past six quarters, Anthropic's revenue has increased more than 45 times, but computing power infrastructure costs have only increased about 10 times. The company has increased revenue per dollar of computing power by supporting multiple chips and improving training algorithms. This is also a key reason for its rapid increase in gross margin. 9. Future goal: A profit margin of over 35%. Anthropic aims for a long-term adjusted operating profit margin of over 35% and to gradually control computing power and infrastructure expenditures to 45%–50% of revenue. However, the company has also signed long-term procurement and leasing commitments totaling approximately $900 billion, the vast majority of which are related to computing power. These payments will be made gradually over the next few years. 10. 950 AI Agents Discover New Biological Structures in 21 Hours In the summer of 2026, approximately 950 Claude Agents spent 21 hours analyzing 1.9 billion protein clusters. One of the agents discovered a previously unpublished CRISPR-like repetitive sequence structure, which was subsequently verified by scientists in their labs. Anthropic believes that the next wave of AI breakthroughs will expand from programming and cybersecurity to the life sciences, mathematics, and robotics fields.