RQ / MARKET METHODOLOGY

A public price.
An open definition.

The Requenta index is planned as a public signal for GPU compute, defined separately for each GPU specification and region. The first planned market is 8×H100 SXM in the European Union.

No fills. No price.

There is no live reserve or trading history yet. Capacity, prices, depth and booking rates are therefore unpublished. The public endpoint returns null for unknown values, rather than treating them as zero or showing illustrative prices.

What the index will measure

The index is intended to represent the price at which eligible compute clears in a market window. A fill means an accepted allocation of a specific node and delivery window at an agreed price. Quotes and expressions of interest will not count as fills.

On-demand = 100

The proposed comparison expresses cleared prices relative to a published, like-for-like on-demand reference. An index of 100 means parity with that reference. The provider source, hardware match, billing basis and currency treatment must be published before this baseline is activated.

What remains to be finalised

The exact aggregation rule, minimum fill threshold, stale-data treatment, reference provider and correction policy will be specified before the first live publication. This draft does not claim that a trading or benchmark methodology has already been validated.

Update cadence

Once the market opens, the planned update interval is 15 minutes. Each published observation will carry its timestamp, SKU, region and relevant availability fields.

Use the data

The pre-launch state is available as JSON at /api/index/latest, cached for 60 seconds. Attribute reused data to Requenta and link to this definition. An embeddable live widget is planned once verified market data exists.