Stacked area
Hyperscaler Capital Spending
Capital expenditures, trailing 12 months, $ billions
Plays the chart through its history. Pauses are where the data set a record or moved most.
From each company’s SEC filings. Includes all capital spending, most of it data centers and servers.
What the four largest cloud companies spend on buildings and equipment, most of it now for AI data centers. Stacked, the layers add up to their combined spending.
What this chart reads
- Amazon capex (trailing 12 months)as “Amazon”latest 2026-06-30
- Microsoft capex (trailing 12 months)as “Microsoft”latest 2026-06-30
- Alphabet capex (trailing 12 months)as “Alphabet”latest 2026-06-30
- Meta capex (trailing 12 months)as “Meta”latest 2026-06-30
Sources
Amazon capex (trailing 12 months). Source: Palanor reads XBRL filings from SEC EDGAR, a public-domain repository. The series uses the PaymentsToAcquirePropertyPlantAndEquipment or PaymentsToAcquireProductiveAssets element that Amazon files under US GAAP.
Microsoft capex (trailing 12 months). Source: Palanor reads XBRL filings from SEC EDGAR, a public-domain repository. The series draws on cash purchases of property and equipment (us-gaap PaymentsToAcquirePropertyPlantAndEquipment or PaymentsToAcquireProductiveAssets) from Microsoft's 10-Q and 10-K filings.
Alphabet capex (trailing 12 months). Source: Palanor reads XBRL filings from SEC EDGAR (public domain). The series comes from cash-flow statements in Alphabet's 10-Q and 10-K filings.
Meta capex (trailing 12 months). Source: Palanor reads quarterly and annual XBRL filings from SEC EDGAR, extracting cash purchases of property and equipment from the cash-flow statement. Data sync weekly via automated process.
Charts are drawn from Palanor's signal library and recomputed from source data on every view. Transformations (changes, indexing, year-on-year rates) are Palanor's.