IPO filing: Anthropic routed 47% of its sales, or ~$2.16B, in 2025 through cloud partners Amazon and Google; analysis: it paid ~$351M back in distribution fees
Reuters:NEW
Context & Ripple Effects
Anthropic enters its IPO process with rapid reported revenue growth but heavy losses: its prospectus lists roughly $4.6B of 2025 revenue alongside a $42B net loss. The new channel disclosure shows that a large share of that revenue reached customers through cloud providers rather than a wholly owned sales path.
The distribution dependence sits alongside more than $518B of planned AI-infrastructure commitments with six partners. Amazon already had an $8B investment commitment and valued its Anthropic stake at $14B at the end of 2024, tying one cloud partner to both Anthropic’s capital base and route to market.
First-order effects
- Amazon and Google receive a direct economic role in Anthropic’s enterprise sales: 47% of 2025 sales flowed through their cloud channels, with about $351M paid in distribution fees.
- Anthropic retains less of the revenue booked through those channels, adding a sales-distribution cost to a business that reported substantial operating losses.
Second-order effects
- Anthropic’s infrastructure partners gain leverage across both customer access and compute supply, particularly given its long-term partner spending commitments.
- For enterprise buyers, Amazon and Google become important procurement gateways for Anthropic models, making cloud marketplace terms and availability part of the vendor choice.
Third-order effects
- If leading model developers continue to rely on cloud marketplaces for distribution while buying infrastructure from the same providers, hyperscalers can capture tolls on both sides of AI-model commercialization.
- The arrangement points toward AI vendors operating as platform-dependent suppliers, with customer ownership and margin allocation increasingly shaped by cloud partners rather than model providers alone.
The trend: AI model companies are trading direct control of distribution economics for the customer reach and infrastructure access of hyperscale cloud platforms.