Learn · track 4 of 4
Reading our data
Every number on Compute Terminal comes from a stored dataset with a source, a date and a method, and every method is public. This track explains how to read them.
It starts with what an H100 hour costs in 2026, read live from the weekly price history. It then shows how to read the price board itself: the three provider classes, the number of providers behind each cell, and the label that says whether a cell is published, indicative or a census of the four hyperscalers. On the latest board the neocloud cells read $3.89 per GPU-hour
Compute Terminal price board (public rate cards and marketplaces)As of 11 Nov 2026Source$4.86 per GPU-hour
Compute Terminal price board (public rate cards and marketplaces)As of 11 Nov 2026Source$7.85 per GPU-hour
Compute Terminal price board (public rate cards and marketplaces)As of 11 Nov 2026Source$8.37 per GPU-hour
Compute Terminal price board (public rate cards and marketplaces)As of 11 Nov 2026Source
The last pieces explain our indices in plain words: the CTI Compute Cost Index, built from those prices; the CTI AI-Factory Equity Index of the listed companies behind the build-out; and the CTI AI-Factory Capacity Index, which follows the power of AI datacentres in operation and in the pipeline. The full rules for each are on the methodology page.
Sources
- Compute Terminal price board, from public rate cards and marketplaces; methodology.
- Compute Terminal Index (CTI) methodology; methodology.
- CTI AI-Factory Equity Index methodology; methodology.
1 · Live figures, updated daily
H100 price per hour in 2026: what a GPU-hour costs now
The median advertised on-demand price of an H100 GPU-hour was $3.93 in the week of 9 Nov 2026, across 44 providers — up 31% on the same week a year earlier ($2.99). The B200 moved +48% over the same year, to $8.14.
2 · 6 min read
How to read the price board
What each cell on the price board means: the provider classes, the sample size, the published, indicative and census labels, and what each plan sees.
3 · 6 min read
The CTI in plain words: the Compute Cost and Capacity indices
What the Compute Terminal Index measures, how the Compute Cost Index turns daily GPU prices into one series, and how the Capacity Index tracks AI datacentre power.
4 · 6 min read
The CTI AI-Factory Equity Index in plain words
How the CTI AI-Factory Equity Index chooses, weights and calculates the listed companies that build and run AI factories, and who oversees it.
5 · 6 min read
The CTI AI-Factory Capacity Index
How the Capacity Index measures the AI datacentre power switched on each week, what goes into its pipeline, and why a revision lands as a step.
Next: the AI Infrastructure Primer 2026 → · Glossary · Start here
Free · every trading day · 08:00 ET