Bloomberg reported that Anthropic is in talks with Google for a long-term cloud computing partnership valued at "multibillion dollars," centered around Google providing increased computing resources, including the use of Tensor Processing Units (TPUs) to accelerate training and inference. The report indicated that the deal has not yet been finalized, and neither party has officially commented. Reuters and other media outlets subsequently reported that Alphabet's stock price rose after the announcement, with the market interpreting this as an increase in Google Cloud's share of high-end AI customers.
Regarding speculation about whether rate limiting will be lifted, there's currently no public documentation directly linking potential hashrate protocols to specific user-side rate limits. Rate limiting strategies are typically governed by factors such as model security, cost control, capacity scheduling, and abuse risk. Even with increased hashrate, product-level request limits and pricing policies will still be subject to official Anthropic announcements.
Frequently Asked Questions
Q: Has the deal been signed?
A: Not yet. Bloomberg News described the deal as "under negotiation," but neither party has released a formal statement or detailed terms.
Q: What are the main contents of the cooperation?
A: Provide Anthropic with larger-scale Google Cloud/TPU and other computing power to support model training and deployment. The term and amount have not been disclosed.
Q: Does this conflict with the cooperation with Amazon?
A: Anthropic maintains in-depth cooperation with Amazon and has received investment from the latter; if it reaches an agreement with Google, it will form a multi-cloud and multi-chip supply pattern.
Q: Will the Pro/Max current limit on the user side be released immediately?
A: There's no definitive answer. The current limit is affected by many factors, so we'll need to wait for Anthropic's product and pricing updates.
Q: Why did the capital market respond positively?
A: The market interprets this as Google Cloud's share of high-value AI workloads may increase, which will benefit the competitiveness of its cloud business.
