Anthropic’s roster of backers reads like a map of tech power: Amazon, Google, Microsoft, and NVIDIA are all on it. This piece covers who put up the money, who supplies the compute, and why those are often the same companies. Purely informational—no directional views on any listed stock. Background: What kind of company is Anthropic.
In a single line: Anthropic’s biggest backers are often its biggest compute partners, so every partnership carries both support and dependence.
The Backers Are Often the Compute Partners
Anthropic’s investor structure has one striking feature: the money and the compute come from the same companies.
The scarcest resource in frontier AI is compute, held by cloud giants Amazon, Google, and Microsoft. By investing, they buy two things at once: a share in Anthropic’s growth, and a heavyweight customer locked into their cloud. For Anthropic, this means funding and stable compute supply on one side, dependence on the other. How this compute line works: Anthropic’s compute gamble.
The Cloud and Chip Giants
The roles of the core investors break down as follows:
- Amazon: Anthropic’s largest investor. Officially confirmed cumulative investment stands at roughly $8 billion, with an additional $5 billion in 2026 and even larger commitments to follow; AWS is also Anthropic’s primary training cloud, running on its in-house Trainium chips. Amazon has clearly stated that its stake is a minority position.
- Google: One of the early investors, and also a provider of TPU compute. However, Google’s exact investment figure has never been fully confirmed officially, and reported numbers vary widely—from billions of dollars early on to Reuters’ 2026 report of a “plan to substantially increase its commitment.” This figure remains pending official confirmation and shouldn’t be treated as settled.
- Microsoft and NVIDIA: In late 2025, the two reached strategic partnerships with Anthropic, with Anthropic committing to purchase a large volume of Azure compute, and NVIDIA and Microsoft each making investment commitments of their own.
All four are publicly listed; this piece states partnership facts only.
Series H: An Expanded Roster of Investors
The Series H round in May 2026 stretched Anthropic’s shareholder list out by another notch.
This round of roughly $65 billion was led by large investment institutions such as Altimeter, Dragoneer, Greenoaks, and Sequoia, with Capital Group, Coatue, D1, and others also participating. Notably, even the three major memory makers—Samsung, SK hynix, and Micron—joined as strategic infrastructure partners.
The inclusion of memory makers is a telling signal: it shows that the supply of HBM (high-bandwidth memory) has come to be viewed by Anthropic as a critical resource worth locking in ahead of time—so much so that it simply turned suppliers into shareholders. The full picture of this funding and valuation is laid out in Anthropic’s valuation and IPO.
Laying out the largest recent rounds side by side makes it clearer just how steep this growth curve has been:
| Round | Date | Amount Raised | Post-Money Valuation | Lead / Main Investors |
|---|---|---|---|---|
| Series A | 2021 | About $124 million | Early; not disclosed | Spark Capital and others |
| Series F | September 2025 | About $13 billion | About $183 billion | Multiple institutions |
| Series G | February 2026 | About $30 billion | About $380 billion | GIC, Coatue |
| Series H | May 2026 | About $65 billion | About $965 billion | Altimeter, Dragoneer, Greenoaks, Sequoia |
(There were also multiple rounds from B through E in between; only the largest recent milestones are picked out here. Valuations are post-money and reflect private primary-market figures. Total funding across 9 rounds: roughly $87.1 billion.)
More Than Money: Enterprise Services and Integration Partners
Anthropic’s partnerships don’t stop at the “take money, rent compute” layer.
In early 2026, it formed a new enterprise AI services company as a joint venture with Blackstone, Hellman & Friedman, and Goldman Sachs, aiming to embed Claude into the core business processes of mid-sized enterprises. The point of this strategy is to cover the last mile of “the model is powerful, but enterprises don’t know how to use it.”
On the implementation side, it also partners with large enterprise-services and consulting firms such as KPMG, Cognizant, Infosys, and IBM, letting them help push Claude into vast enterprise customer bases. Add to that the open MCP (Model Context Protocol) ecosystem, which makes it easier for Claude to connect with external tools, and Anthropic is effectively extending from “selling a model” toward “weaving an enterprise ecosystem.”
The Two Sides of Support and Dependence
Anthropic has pulled nearly every heavyweight player onto its team: the largest cloud provider, the largest chipmaker, and the major memory makers are all shareholders. Multiple layers of insurance across funding, compute, and supply chain.
The risk side: when backers are also suppliers and potential competitors, every contract shifts a delicate balance of power. Anthropic enjoys the backing but must manage the dependence.
Penchan’s Take
Frontier AI burns too much cash for any company to go it alone. Anthropic’s solution: turn the cloud and chip giants into shareholders. Deep binding for security — but the same binding ties its fate to these giants. Watching how these relationships shift is more revealing than tracking the dollar amounts.
Further reading: What kind of company is Anthropic, Anthropic’s compute gamble, Anthropic’s valuation and IPO.