Anthropic is approaching its IPO in an unexpectedly contradictory position. On the one hand, the company continues to demonstrate rapid revenue growth and has achieved adjusted operating profitability for the first time. On the other hand, corporate clients are increasingly moving away from its most expensive flagship model, Fable 5, in favor of more affordable alternatives. This is an important signal for the market, as companies are beginning to evaluate AI based not only on quality but also on the economics of using it.
Demand for Fable 5 among corporate clients has virtually stopped growing. An increasing number of companies believe that the capabilities of earlier models are sufficient for most business tasks and that paying a premium for the most advanced system is not always justified. Against this backdrop, Anthropic’s own cheaper model, Opus 5, has already surpassed Fable 5 in corporate spending.
Competition is also intensifying. After the launch of GPT 5.5 in July, OpenAI significantly strengthened its position, and its projected annual revenue exceeded $40 billion. At the same time, the new model is significantly cheaper than Fable 5. Additional pressure comes from open models from China and other countries, which allow businesses to cut costs even further.
This is particularly important ahead of the Anthropic IPO, which may take place as early as September or October. Investors are discussing a potential market capitalization of $2 trillion and an offering that could surpass one of the current stock screener leaders, SpaceX, in terms of the amount of capital raised. However, to justify such a valuation and support Anthropic stock after its market debut, the company needs to prove that its growth is not dependent solely on sales of its most expensive models.
For now, the financial momentum remains strong. In July, Anthropic reported that its annualized revenue had reached $65 billion, up from $47 billion in May, and since the beginning of the year, the figure has grown nearly sevenfold. In the second quarter, the company became profitable on an adjusted operating basis for the first time and expects to remain profitable in the third.
Nevertheless, investors are beginning to be concerned about the changing structure of demand. If corporate clients are indeed increasingly opting for mid range models, this could change the economics of the entire AI industry. Until now, developers have justified the multibillion dollar costs of training new systems by arguing that the most powerful models would generate the bulk of future revenue. Now it increasingly looks like flagship models may become more of a demonstration of technological leadership, while the real business is built around cheaper and more efficient solutions.
Additional risks come from the infrastructure required to support this growth. Anthropic plans to state explicitly in its IPO prospectus that public opposition to the construction of data centers could negatively affect its business development. Only about a quarter of Americans support the construction of AI data centers near their homes, while approximately 70% oppose it.
For Anthropic, this is not just a reputational issue. Restrictions on data center construction could raise infrastructure costs and slow the deployment of new computing capacity, without which the company may struggle to maintain its current growth rate. At the same time, energy is already becoming one of the most expensive components of new AI projects.
This problem is compounded by the rising cost of the accelerators themselves. Nvidia has already notified customers that, starting next year, certain server systems based on Vera Rubin and Grace Blackwell will increase in price by more than 15% due to rising memory prices. At the same time, Nvidia stock came under pressure, recording its longest losing streak since September 2022 ahead of the company’s earnings report.
Thus, pressure is coming from two directions at once. On the one hand, Anthropic’s customers are increasingly demanding cheaper AI models. On the other hand, the cost of computing infrastructure continues to rise. As a result, AI developers have to find a balance between technological leadership and profitability.
This question is likely to become one of the key considerations for investors assessing Anthropic ahead of its IPO. The company has already proven that it is capable of rapid revenue growth, but now the market needs to understand how sustainable its profitability will be in a scenario where customers are shifting toward cheaper models, infrastructure costs are rising, and competition is intensifying. If this trend continues, the main advantage in the next stage of the AI race may no longer be having the most powerful model, but the ability to deliver sufficient quality at the lowest possible cost.
This content is brought to you by the FingerLakes1.com Team. Support our mission by visiting www.patreon.com/fl1 or learn how you send us your local content here.