SCORING
Scoring
Composite quality on a fixed 30 / 30 / 25 / 15 workload blend across reasoning, code, extraction, and classification. Value score uses blended $/M divided by composite quality / 100.
The TKX grounds quality to cost on a published method. Baseline 100 = 2026-08-04. Current readings are on the live index.
A TKX print above 100 means a unit of AI capability is at a premium to the 2026-08-04 baseline of 100. A print below 100 means it is lower than at that baseline.
As of 2026-09-13 · Source: Cambent TKX
Composite quality on a fixed 30 / 30 / 25 / 15 workload blend across reasoning, code, extraction, and classification. Value score uses blended $/M divided by composite quality / 100.
Construction v0.3 chains the printed TKX, Open, and Frontier levels when a fully scored general is added or dropped, so a listing does not jump the number; the composition effect is shown separately that day.
We publish what we can source for the exact model you buy. When we can’t yet, we mark it plainly — Grounded, Partially grounded, or Provisional — and keep grounding more. An honest gap beats a silent guess.
Every quality score is labeled Grounded, Partially grounded, or Provisional. Grounded means published benchmarks and a sourced price for that exact model. We show the source and date. Partially grounded means some scores for this exact model are from a named source; others are not published yet. We say which. Missing stay unlabeled. Never estimated from a close version. Provisional for list-only means listed, not scored yet. Other provisional estimates say grounding is in progress without inventing a score. Machine-readable coverage is on the Index API.
Scoring recipe dated 2026-08 (v0.2). Construction chaining dated 2026-08 (v0.3).