Anthropic announced a 4.2 billion dollar Series G round on July 11 2026. Amazon contributed 1.8 billion and Google added 1.4 billion with the remainder from existing investors. The round closed at a 75 billion post-money valuation.
Proceeds will purchase 200,000 Nvidia H200 GPUs for Claude 4 training. The new cluster will come online in Oregon by December 2026. Anthropic also plans to open three additional safety research offices in London, Singapore, and Toronto.
The company previously raised 8.5 billion dollars across five rounds since 2023. Claude 3.5 Sonnet launched in January 2026 and currently powers 38 percent of enterprise AI copilots. Revenue reached 1.1 billion dollars annualized in June 2026.
Amazon holds a 28 percent stake while Google owns 19 percent after this round. Both cloud providers receive preferred access to future model weights. Anthropic maintains an independent safety board with veto power over releases.
Why this matters
The valuation sets a new benchmark for AI startups and intensifies competition for compute resources. Smaller labs may struggle to match training scale. Policy makers will scrutinize the Amazon and Google stakes for potential antitrust concerns.
Anthropic expects Claude 4 to ship in October 2026 with improved agentic capabilities. The company will publish updated constitutional AI research alongside the launch.