# Who does Anthropic pay? Who are its key suppliers?
Canonical: https://valuechain.wiki/anthropic · Facet: suppliers · Updated: 2026-09-04T23:10:42.679484+00:00 · Source: ValueChain.wiki (cited money-flow graph)

**Citable summary**: Anthropic's confirmed suppliers/cost counterparties include 아마존, 알파벳 (구글), Broadcom plus 2 undisclosed/aggregate relationships. Per the valuechain.wiki graph: 5 cost edges, backed by 8 dated sources, 0 of them primary filings (as of 2026-09-04).

### 아마존 (high)  → /amazon.md
- note: Uses 5GW of AWS Trainium; Amazon invests up to $25B ($100B AWS commitment).
- [confirms] Anthropic (EN, ir, published 2026-04-19) https://www.anthropic.com/news/anthropic-amazon-compute: "Amazon invests up to $25B; Anthropic commits $100B to AWS over 10 years."
- [confirms] SemiAnalysis (EN, research, published 2025-09-03) https://newsletter.semianalysis.com/p/amazons-ai-resurgence-aws-anthropics-multi-gigawatt-trainium-expansion: "Three AWS campuses (>1.3GW) built solely for Anthropic's training; nearly a million Trainium2 in the largest."

### 알파벳 (구글) (high)  → /alphabet.md
- note: Uses 5GW of Google Cloud TPUs (1M); Google invests up to $40B.
- [confirms] TechCrunch (EN, news, published 2026-04-24) https://techcrunch.com/2026/04/24/google-to-invest-up-to-40b-in-anthropic-in-cash-and-compute/: "Google invests up to $40B in Anthropic, securing 1M TPUs / 5GW."
- [confirms] Anthropic (EN, ir, published 2026-04-21) https://www.anthropic.com/news/google-broadcom-partnership-compute: "Google invests up to $40B in Anthropic, securing 1M TPUs / 5GW."
- [confirms] Futurum Group (EN, research, published 2026-04-09) https://futurumgroup.com/insights/anthropics-gigawatt-scale-tpu-deal-with-broadcom-creates-a-structural-advantage/: "Signed a multi-gigawatt TPU compute deal with Google and Broadcom, capacity from 2027."
- [confirms] Global Economic (KO, news, published 2026-02-19) https://www.g-enews.com/article/Global-Biz/2026/02/202602191747342709fbbec65dfb_1: "Anthropic relies on Google/Amazon/MS cloud for Claude; Google invests heavily."
- [confirms] Wikipedia (KO, news, published 2026-01-01) https://ko.wikipedia.org/wiki/앤트로픽: "Anthropic relies on Google/Amazon/MS cloud for Claude; Google invests heavily."
- [mentions] SemiAnalysis (EN, research, published 2025-09-03) https://newsletter.semianalysis.com/p/amazons-ai-resurgence-aws-anthropics-multi-gigawatt-trainium-expansion: "A large share of Anthropic's spending goes to Google Cloud."

### Broadcom (mid)  → /broadcom.md
- note: Expands custom compute via Broadcom, the Google-TPU design partner.
- [confirms] Anthropic (EN, ir, published 2026-04-21) https://www.anthropic.com/news/google-broadcom-partnership-compute: "Anthropic-Google-Broadcom partnership for multi-GW next-gen compute."
- [confirms] Futurum Group (EN, research, published 2026-04-09) https://futurumgroup.com/insights/anthropics-gigawatt-scale-tpu-deal-with-broadcom-creates-a-structural-advantage/: "Broadcom supplies custom AI chips and optical interconnects for Google TPUs; the integration layer."

### NVIDIA (mid)  → /nvidia.md
- note: Uses NVIDIA GPUs as a secondary training/inference substrate.
- [confirms] Anthropic (EN, ir, published 2026-04-21) https://www.anthropic.com/news/google-broadcom-partnership-compute: "Claude runs across TPU, Trainium and NVIDIA GPUs."
- [mentions] Futurum Group (EN, research, published 2026-04-09) https://futurumgroup.com/insights/anthropics-gigawatt-scale-tpu-deal-with-broadcom-creates-a-structural-advantage/: "Runs Claude across AWS Trainium, Google TPUs and NVIDIA GPUs."

### CoreWeave (mid)  → /coreweave.md
- note: Rents GPU-cloud compute from CoreWeave.
- [confirms] ZDNet Korea (KO, news, published 2025-09-16) https://zdnet.co.kr/view/?no=20250916175927: "Rents GPU-cloud compute from CoreWeave."

Reverse side: https://valuechain.wiki/anthropic/customers.md · Ripple: https://valuechain.wiki/anthropic/ripple.md

---
Full node: https://valuechain.wiki/anthropic.md · JSON: https://valuechain.wiki/anthropic.json