Anthropic Locks In $65B From Amazon and Google — And 10 Gigawatts of Compute
Google committed up to $40B and Amazon up to $25B in Anthropic over days in late April, lifting its valuation to $350B and securing 10 GW of dedicated hyperscaler compute. It's the fastest accumulation of strategic capital ever directed at a single private tech company.
Original sourceAnthropic closed the largest and fastest funding round in tech history during the final week of April 2026. Google announced an investment of up to $40 billion on April 24 — $10 billion immediately at a $350 billion post-money valuation, with up to $30 billion more contingent on performance targets — following Amazon's announcement of an additional $25 billion commitment days earlier.
Combined, Anthropic now has roughly $65 billion in fresh hyperscaler commitments and, critically, pledges of 10 gigawatts of AI compute: 5 GW on Google's TPU infrastructure and 5 GW reserved through Amazon Web Services. The Google deal specifically locks in dedicated TPU compute capacity, not just credits — meaning Anthropic has guaranteed training and inference throughput independent of the public cloud spot market.
The $350 billion valuation makes Anthropic the second-most-valuable private company in the world after SpaceX, and it represents a compression of the fundraising timeline that's historically unprecedented. OpenAI took several years to raise equivalent commitments; Anthropic compressed an equivalent escalation into roughly twelve months.
The strategic logic for both hyperscalers is defensive: each wants Anthropic's models deeply integrated into their cloud platforms to compete with the other's AI offerings. Google's TPU deal ensures Anthropic trains future Claude versions on Google infrastructure, while Amazon's AWS deal ensures Claude remains the anchor model in Bedrock. Anthropic gets the compute it needs to train frontier models; the hyperscalers get a credible frontier-model partner that isn't OpenAI.
For the AI industry broadly, the deal signals that hyperscalers view foundation model companies as critical infrastructure worth securing at essentially any cost — and that the race to lock up the top-three frontier labs (OpenAI, Anthropic, DeepSeek/domestic) as strategic partners is now essentially complete.
Panel Takes
The Builder
Developer Perspective
“10 GW of guaranteed compute means Claude models aren't going to hit capacity walls during training runs the way GPT-4 did in 2023. For developers building on the Claude API, this is the strongest signal yet that Anthropic will be around and capable of competing for the next decade.”
The Skeptic
Reality Check
“A $350B valuation for a company with roughly $1.5B in annual revenue implies a 230× revenue multiple. That's not an investment thesis, it's a geopolitical hedge. When the hyperscaler AI spending supercycle ends, Anthropic's implied valuation will look very different — and the $65B could evaporate into capex with no return.”
The Futurist
Big Picture
“The fact that both Google and Amazon are paying up to avoid letting the other have Anthropic exclusively tells you everything about where they think AI is going. This is the hyperscaler bet that foundation models become the OS layer of the next computing platform — and they're paying platform-company prices to secure that position.”