STCI · Standard Token Cost Index

The benchmark price for AI tokens.

Oil has Brent and WTI. Shipping has the Baltic Dry Index. Tokens are becoming a commodity too, and they have no benchmark that the sellers do not control. STCI is a public, reproducible start on one.

1. Tokens need a benchmark price

Every commodity market that works has a reference price that no single seller sets. Brent and WTI price crude. The Baltic Dry Index prices bulk shipping capacity. Contracts, budgets and hedges are written against those numbers.

Tokens have none. Each provider sets its own price, changes it without notice, and more and more tokens are subsidized or resold by intermediaries. A buyer cannot tell a fair price from a subsidy, budget a year ahead, or hedge.

2. What a benchmark needs

Routes

Standard units everyone prices the same way. Model tiers today; the cost of a defined task later.

A panel

Gateways and large buyers reporting what they actually paid, anonymized and weighted by volume.

A public rulebook

How every value is computed, written down before it is used.

An audit trail

Values anyone can reproduce, and administration independent of the sellers.

3. Not every token is the same token

Crude has grades: light, heavy, sweet, sour. Tokens have grades too: serving precision, context length, uptime. The neocloud spread is the first measurement of that.

The spread measurement publishes with methodology 2.0.0. No spread file is in the published data yet.

4. Where STCI is today

  1. Now Listed prices. Every separately priced model on one aggregator (OpenRouter), normalized to $ per million tokens, combined into four indices under a public rulebook. Each day's value carries a verification hash that anyone holding the inputs can reproduce.
  2. Now Grades and spread. For open-weight models, the price of every provider endpoint selling the same model, and the precision each one serves.
  3. Next A reported-price panel. Gateways and large buyers report what they actually paid, anonymized and volume-weighted. Listed prices cannot show discounts, credits or subsidies; a panel can.
  4. Later Contracts that settle against the index. Forwards that let a buyer fix next year's token cost. This needs the panel and independent administration first.

5. Today's numbers

Blended $ per million tokens at 3 output tokens per input token, for March 6, 2026 (methodology 1.0.0).

STCI-FRONTIER

$8.67 per 1M tokens

Flagship models, one per lab. 9 models.

input $2.17 · output $10.83

STCI-EFFICIENT

$1.75 per 1M tokens

Cost-optimized models, one per lab. 8 models.

input $0.51 · output $2.16

STCI-OPEN

$1.94 per 1M tokens

Open-weight models sold by many providers. 7 models.

input $1.07 · output $2.23

STCI-ALL

$6.69 per 1M tokens

Every eligible listed model. 324 models.

input $2.02 · output $8.25

The neocloud spread

Source: listed prices on OpenRouter, the model catalog and each model's provider endpoint list. Listed prices, not prices paid. JSON: /v1/index/latest.json, /v1/spread/latest.json.

6. Report what you pay

The next rung needs prices paid, not prices listed. If you run an LLM gateway, or your company buys tokens at volume, you can make the index real by reporting what you pay, anonymized and aggregated before anything is published.

Email jeremy@intentsolutions.io The code and data on GitHub

7. Methodology and reproducibility

The rulebook is public. Today's value carries verification hash c3dee53e17e24e60: SHA-256 over the sorted eligible observations, the methodology version and the date. Anyone holding the inputs can recompute it; the observations are available on request.

Series coverage, labeled as it is:

Values across the break are not comparable: the baskets changed with methodology 2.0.0. Methodology and API.