Summary of Key Points
On May 29th, AI large-model company Anthropic completed a $65 billion Series H financing, raising its post-financing valuation to $96.5 billion, surpassing OpenAI’s previous valuation of $85.2 billion and becoming the most valuable large-model startup in the world. Additionally, Anthropic released the more powerful new model Claude 4.8, which saw its annual revenue soar from $14 billion to $47 billion within just three months. By partnering with memory chip giants and cloud service providers, it has addressed the issue of computational power shortages. In contrast, OpenAI’s progress has slowed down recently, and its planned IPO seems passive, pushing the competition between the two companies to a fever pitch before they go public.
Detailed Analysis
#### 1. Valuation Surpass: Why Has Anthropic Become the New Favorite of Investors?
Anthropic’s valuation now exceeds that of OpenAI, indicating a shift in the leadership within the AI community. The reason investors are willing to invest heavily is due to its recent remarkable progress: continuous breakthroughs in model technology, rapid growth in corporate clients, and an advanced approach to computational power infrastructure. OpenAI has not made significant moves recently, and even its planned secret IPO submission seems to be falling behind Anthropic’s pace. This suggests that investors are more optimistic about Anthropic’s future growth potential, as in the AI industry, the company that can deliver better products and acquire more customers faster will become the most valuable.
#### 2. The Capital Alliance Goes Beyond Just Funding; It Solves the Computational Power Challenge
The list of investors in this round is particularly noteworthy, including traditional giants like Sequoia and GIC, as well as global memory chip leaders Micron, Samsung, and SK Hynix. These partners are not just providing funding but also acting as strategic collaborators for Anthropic. The biggest bottleneck for AI large models is computational power, especially high-speed memory (HBM). These chip companies can ensure that Anthropic has a steady supply of HBM, which is essential for its computing clusters (used in collaborations with Amazon, Google, and SpaceX). While OpenAI is also making similar arrangements, Anthropic has already secured support from three major cloud platforms, putting it at an advantage.
#### 3. The New Model Claude 4.8: World-Leading in Programming and Operating System Capabilities without Price Hikes
Anthropic’s newly released Claude 4.8 Opus is a significant technological highlight ahead of its IPO. This model outperformed others in two key tests:
- Programming Tasks: It achieved an accuracy rate of 69.2% in the SWE-bench Pro test, which evaluates AI’s ability to write code independently to solve engineering problems, surpassing both GPT-5.5 and Google Gemini.
- Operating System Interaction: It scored 83.4% in the OSWorld test, which simulates real computer operations such as file management and software installation, also ranking first.
More importantly, Claude 4.8 introduces a new approach to AI interaction through dynamic workflows, allowing multiple AI entities to collaborate on complex tasks like a team. The model’s cost remains affordable at $5 per million input tokens, making it more attractive to corporate clients.
#### 4. Revenue Tripling in Three Months: Corporate Clients Are Flocking to Use Its Services
Anthropic’s annual revenue has increased by more than three times, from $14 billion at the beginning of the year to over $47 billion, demonstrating high customer acceptance. Many companies use its models for coding, contract processing, and business process optimization, willing to pay for these services. This rapid revenue growth supports further investment in research and development as well as computational power, creating a positive cycle.
#### 5. The Race Before Going Public: Anthropic Leads, OpenAI Lags Behind
Both companies are vying to be the first to list as the “number one AI large-model company.” Anthropic currently has an edge in terms of technology (model performance), capital (higher valuation), and revenue growth. Although OpenAI is more well-known among general users (e.g., with ChatGPT), it has appeared slower in the corporate market and technological updates. Experts suggest that the AI industry is highly competitive, and leadership can change quickly. The outcome of this race will not only determine which company goes public first but also shape the future of the entire AI sector.
Conclusion
Anthropic’s recent financing and model updates mark a new phase in the competition for AI large models. By securing strategic partnerships and solving computational power issues, it has attracted customers with its technological advancements. Its revenue growth validates its commercial value. If OpenAI cannot accelerate its progress, it may lose its leading position. For users, this competition will result in more user-friendly and affordable AI products; for the industry, it will drive faster adoption of AI technologies across various fields.