Apollo and Blackstone Reportedly Back $35 Billion Anthropic Chip Financing as Deal Details Remain Unclear

This transaction has drawn attention first and foremost because of its enormous size. For a company like Anthropic, which sits at the center of competition in AI infrastructure and large models, financing at this scale is usually more than routine operational fundraising. It is more likely tied to compute procurement, data centers, power capacity lock-ins, chip supply, or long-term infrastructure commitments. The involvement of large alternative asset managers such as Apollo and Blackstone also makes the market more inclined to view this kind of deal as a continuation of the “capitalization of AI infrastructure,” rather than simply a traditional private equity investment.

However, different sources describe the situation differently, and the information still awaits official confirmation. Existing RAG information shows that Anthropic had previously been reported to be in talks for at least $30 billion in new financing, with a focus on expanding computing infrastructure. But this background does not directly confirm that the current “$35 billion chip project financing” has been completed, nor does it confirm whether it is the same transaction or financing vehicle as the previously reported company-level fundraising. Before more details are disclosed, the market still needs to distinguish whether this is Anthropic corporate financing, financing for a specific compute/chip SPV, or a project-level arrangement tied to infrastructure partners.

From a market-context perspective, if this financing is ultimately confirmed, the signal extends beyond AI venture capital and into a broader technology capital expenditure cycle: whether top asset managers are beginning to treat AI compute, chips, and data centers as new long-term allocable assets. For the crypto market, while this is not a direct event, the continued absorption of large-scale institutional capital by AI infrastructure will also affect the market’s valuation framework for the “AI narrative,” the financialization of compute assets, and AI-Crypto crossover sectors.

## Why It Matters

The key point of this news is not Anthropic itself, but the potential shift in capital structure. If it is ultimately confirmed that large alternative asset managers such as Apollo and Blackstone are leading or deeply participating, it would suggest that AI infrastructure financing is moving away from traditional VC/growth equity logic and toward a model closer to project finance built around power, data centers, equipment leasing, and packaged long-term cash flows.

Its importance to the market lies in the fact that the competitive barrier in AI is shifting further from model capability toward capital intensity and the ability to lock in infrastructure. Whoever can secure long-term funding, chip supply, and power resources faster will have a better chance of sustaining training and inference at scale. But current disclosures remain incomplete, especially since the financing instrument and project boundaries are still unclear, so the market cannot yet judge whether this will become a replicable, standardized financing model.

## WEEX View

The core question is not simply “how much Anthropic raised again,” but whether, if the $35 billion figure is real, this is continued primary equity pushing valuations higher, or old money entering directly into underlying AI assets through debt, preferred capital, or project SPVs. The former is a growth story; the latter is a move to capture future cash flows and control over collateral. For front-line trading businesses, the market implications of these two structures are completely different: if this is equity financing, the AI narrative remains more about valuation spillover; if it is project financing, it suggests traditional large capital has already begun treating compute, power, data centers, and chip contracts as financializable assets, and capital will continue rotating out of high-volatility risk assets into quasi-infrastructure targets backed by underlying contracts.

From a CEX perspective, what matters more is the path of liquidity migration, not short-term sentiment. In crypto, the broader AI narrative has mainly mapped onto AI tokens, DePIN, compute-related themes, and data infrastructure narratives, but the real profit pool is not in secondary-market tokens. It is in upstream hardware, cloud resource scheduling, and long-term supply agreements. If these kinds of deals continue to increase, AI tokens in the secondary market may face a more practical problem: the narrative heat remains, but capital pricing power is being taken by off-chain institutions, and the room for arbitrage will keep narrowing. The next three things to watch are whether officials disclose the financing structure, whether verifiable supporting chip or data center agreements emerge, and whether more PE or credit giants follow with similar transactions. Once this path is proven out, the point of convergence between AI and Crypto may not first emerge at the application layer, but instead in compute rights, packaged income rights, and on-chain mapped assets.