Frequently Asked Questions

Ornn builds financial infrastructure for the GPU compute economy. We provide reference pricing, hedging instruments, and market data so institutions can transact compute the way they transact any other commodity.

Accessing compute is the bottleneck for frontier AI and multi-year compute contracts can cost hundreds of millions of dollars. Today there is no benchmark price, no hedging instrument, and no standard contract.

Three audiences. Institutional traders and operators hedging compute exposure, buyers and lenders that need a defensible reference price, and analysts pricing compute against the market rather than a single vendor’s rate card.

OCPI is the Ornn Compute Price Index, a family of transaction-based indices tracking clearing prices for rented GPU compute in USD per GPU-hour. It uses anonymized executed trades from a diversified set of independent operators. No offers or indicative quotes are included.

H100, H200, A100, B200, RTX 5090, and RTX 6000. We add new GPU types as transaction volume on them crosses our eligibility threshold.

Daily, published at 16:00 EST.

Yes. The H100 index trades under ticker ORNNH100.

Daily settlements are published at index.ornn.com with additional data found at data.ornn.com. Bloomberg Terminal subscribers have access under the ORNNH100 ticker family.

• Aggregate demand from new model architectures and enterprise AI workloads
• Effective supply tied to hardware delivery and power
• Geopolitics and export controls
• Efficiency gains that migrate workloads across hardware generations

A rate card is one provider’s posted price for their own capacity. OCPI is a market-wide settlement built from executed trades across many providers, regions, and contract tenors, and it is the same number every counterparty sees.

Yes. Contact the Ornn team via email to scope a licensing agreement.

Buyers go long at a forward tenor matched to their training or inference plan. Operators sell forward to lock in revenue, and lenders and insurers use OCPI as a mark-to-market reference. Reach out to the Ornn team via email to discuss more.

OTPI is the Ornn Token Price Index, a family of transaction-based indices that track the realized market price of frontier-lab language-model tokens in USD per million tokens, written $/MTok. For each lab, the index is the token-volume-weighted average price actually paid across that lab’s models over a daily UTC window. It is built from same-day transacted prices, provider weights pooled across the day’s intraday routing samples, and same-day token volumes, all drawn from executed on-demand inference traffic. Each day is settled once and never restated.

They price the two sides of the AI economy. OCPI measures the market’s input, the price of GPU-hours, and OTPI measures its output, the price of tokens. OCPI tells you what it costs to rent the compute; OTPI tells you what the market actually pays per token of model output once caching, routing, and model mix are counted. Both settle the same way.

OTPI is organized one index per lab. It currently tracks token usage across OpenAI, Anthropic, and Google models. Each lab index covers that lab’s priced models as listed on the on-demand token exchanges, with models identified from the exchange catalog and assigned to a lab by namespace. New models enter the relevant lab index as they list and begin transacting. Ornn is also experimenting with open source models, which remain outside the published index for now.

Free variants and zero-price models are excluded because they carry no transaction value. Open-weight models such as gpt-oss and Gemma are excluded because their tokens are sold by third-party hosts rather than by the lab, and OTPI measures tokens transacted on the labs’ own priced offerings. Volumes count paid variants only.

Both are already inside the transacted prices. Cache discounts are embedded in the input price, so fresh input, cache reads, and cache writes are blended in the input term. Reasoning tokens bill within completion at the output price, so visible output and reasoning output share the output term. Because the index is built from executed traffic, these cost classes register exactly as buyers incur them, and the split between input and output does not distort the cost.

OTPI weights each provider by its share of the day’s token throughput. That share is built by summing the provider’s hourly throughput samples across the full UTC day rather than reading a single instant. Instantaneous provider totals are volatile, so pooling across the closed day removes that noise and the weight reflects the day’s overall routing.

Daily, on the exchanges’ UTC day boundary. A day closes at 00:00 UTC, which is 8:00 PM Eastern, and settles 12 hours after close, in practice at 12:00 UTC the next day, once upstream counters have finalized. Only fully closed days are computed, and each day is computed once under a single same-day regime of prices, pooled weights, and volumes.

Several forces register in the index. New model launches and their early adoption dynamics move it, as does the mix of demand across premium and budget tiers, including buyers trading up to more capable models or migrating to cheaper ones. Cache-discount economics and the input and output composition of real workloads feed through, and reasoning intensity pushes volume toward higher-priced output tokens. Provider routing across the serving providers for each model matters, and lab-level monetization and rate-card changes register as they appear in executed traffic.

A rate card is one lab’s posted price for its own models. OTPI is a market-wide settlement built from executed trades across every serving provider, blended by the volumes buyers actually transact, and it is the same number every counterparty sees. As a unit-value index it captures cache discounts, tier mix, and model migration that a posted price cannot show.

Two design choices make it hard to move. First, every input is executed, paid inference, so influencing the index means committing capital at market prices. Second, the cross-model volume weighting uses complete daily token totals, on the order of hundreds of billions to trillions of tokens across millions of requests, so it carries no sampling error, and the within-model provider mix is pooled across the day’s hourly samples to suppress single-reading noise. Settling only after counters finalize means published history is stable by construction.

OTPI is currently published on the Ornn Data platform at data.ornn.com.

Limited access to OTPI data is available through index.ornn.com. For a licensed or higher-volume feed, contact the Ornn team.

Yes. Contact the Ornn team by email at wayne@ornn.com to scope a licensing agreement.

OTPI gives you a transparent benchmark for the cost side of any token-consuming product. Use it to budget and forecast token spend, to benchmark the prices you pay against the market, and to compare monetization across labs on a like-for-like basis. As token markets mature, OTPI is built to serve as a settlement-grade reference for contracts written on token cost. The Ornn team can help you scope a specific use.