← Index
Thread
Compute cost
What it costs to build and serve intelligence, and who pays for it. 4 notes so far.
- November 2026MarketsWhen AI models are free, who gets paid?NVIDIA reports on 17 November with no data-centre chip sales to China in its forecast. Yet its own filing says many of the most popular open models are Chinese, and it is paying $11.9bn for the site that distributes them.Everyone else: $16.9bn → $40.3bn a quarter
- November 2026InfrastructureThe neocloud GTM reset: sell the GPU-hour and you give away the engineeringSix providers charge exactly the same to run OpenAI's gpt-oss-120b; the fastest is eight times quicker than the slowest. Yet 98% of CoreWeave's revenue is still rented capacity, and capacity is getting dearer.Same price: 46 → 394 tokens a second
- November 2026MarketsAnthropic's S-1: revenue per compute dollar rose from $0.40 to $2.08Revenue grew 14 times in a year; compute cost grew 3.8 times. The AI race is becoming a capital allocation problem.$0.40 → $2.08, then $1.72 per $1 of compute
- January 2026InfrastructureToken bills grow with every step an agent takes. A cluster you own doesn't.Metered inference costs 2.4¢ a case and owned capacity 8.7¢. Once agents use about 3.6 times the tokens, owning is cheaper, and agents typically use four times.Owning wins past ~3.6× tokens per case