Token Price Cuts Drive Up Compute Demand in a Jevons Paradox
Understand the story
On October 11, 2026, The Decoder, citing an a16z analysis, reported that the AI market is showing a textbook case of the Jevons Paradox: based on data from Ornn, Silicon Data, and Bloomberg as of August 2026, token prices keep falling while H100 GPU rental prices remain stable or even rise. The reason is that cheaper tokens have fueled the growth of AI agents and automation applications, with usage growing faster than unit prices are falling, ultimately driving up overall compute demand. a16z believes this phenomenon vindicates the earlier judgment of NVIDIA CEO Jensen Huang: efficiency gains amplify rather than reduce compute demand. The report is based only on market data from the above institutions as of August 2026, and no further developments have been reported.
AI overview of reporting · Updated 7 days ago
Reporting timeline
Explore different aspects of the story through its reporting.
- The DecoderCheaper AI tokens are driving more demand, and that's Jensen Huang's best-case scenario
a16z says data from Ornn, Silicon Data, and Bloomberg through August 2026 shows a textbook Jevons paradox in the AI market: token prices keep falling, yet H100 GPU rental prices hold steady or rise, as cheaper tokens fuel AI agents and automation with usage growing faster than prices drop.
Story heat trend
Current heat3·Comparable peak10（10/11 19:00）·Comparable change over 24 hours -49%
The trend compares the same continuously observed sources, which may cover fewer sources than the current heat score. Hover or select the chart to inspect each hour; use the left and right arrow keys to navigate.