The compute spot index.
One citable number for GPU spot pricing: a ten-model basket over every feed we read, chained so that coverage changes cannot move it — only prices can. The methodology is published in full below and the raw series is one click away. Free to read, free to audit.
Methodology, in full.
Basket: ten fixed models spanning datacenter, workstation and consumer silicon, each with multi-provider coverage at inception:
Construction: per model, per half-hour snapshot, we take each provider’s median listed $/GPU/hr — medians, so no single strange listing matters. Between adjacent snapshots a model compares only providers present on both sides; the index step is the geometric mean of those model ratios, and levels chain from 100 at inception. A new feed joining the panel therefore cannot move the index — that is the difference between an index and a coverage counter. The headline uses on-demand listings only (firm prices); the same construction over interruptible listings is published as the idle sub-index, volatile by nature. Steps with fewer than five chainable models carry flat and are counted openly in the payload.
Revision policy: the basket and construction are frozen as v1; any change arrives as an announced version bump, never a silent edit.
Why now.
Compute is becoming a traded commodity: CME Group and Silicon Data list the first regulated compute futures on October 5, 2026 — contracts settling on proprietary H100 and B200 rental indexes. This index is the open counterpart: methodology in full view, every input on our public tape, per-model series for the contracts’ own underlyings readable straight from /api/history. When someone asks “what did GPU spot do this month”, the answer shouldn’t live behind a license.
Honesty box.
The index is young — its tape starts 23 Aug 2026, and the page you are reading will still say so when that sentence is embarrassing. It measures listed marketplace prices, not negotiated contracts or hyperscaler rates. An index over listing medians moves when cheap inventory is absorbed as well as when hosts reprice: both are real market signal, and we name both. It also covers the marketplace tier only — the cluster clouds that sell H100s by the rack are a separate series at the cluster tier, kept apart because posted rates and transacted prices are different numbers.
Take it with you.
An index is worth what cites it, so everything here is built to leave. CC BY 4.0 — use the numbers anywhere, commercially, and name where they came from. The methodology is frozen per version, so a figure you quote today means the same thing when someone checks it.
Cite this
compute.pangle.online GPU spot index (v1), compute.pangle.online/spot/, accessed —. CC BY 4.0.
Embed the live index
A self-contained card, no script on your page, no tracking:
<iframe src="https://compute.pangle.online/embed/" width="320" height="150" style="border:0" title="GPU spot index" loading="lazy"></iframe>
Take the data
index as CSV — the whole series, spreadsheet-ready · JSON — both indexes, gap counts and methodology · the raw tape — every half-hourly observation behind it · a live badge for a README.
Subscribe instead
Field Notes, weekly — where the index sat and what moved more than 5%, once a week. Or the daily wire if you want every move.
Governance: methodology, the 30-day notice rule, corrections, cessation and an honest IOSCO self-assessment · where the data comes from and how concentrated it is.
Raw: /api/index (full series, both indexes, gap counts). The tape behind it: /api/history.