OpenAI Opens Decisions API at 2.4x the Price of Rival Jev

OpenAI made its Decisions API generally available on 22 October 2026, at a list price of $0.10 per million input tokens, about 2.4 times the $0.042 that TypeSafe lists for its rival Jev model, by our arithmetic. OpenAI’s announcement says the API helps “applications and agents make near real-time decisions from text, images, and structured inputs up to 10x faster than GPT‑6 Luna through the Responses API.”

TechCrunch reported the launch as OpenAI entering a market that “recently made TypeSafe, a startup founded by a former researcher at the frontier lab, a $7.6 billion company.” As we read it, OpenAI is not competing on list price: of the six decision models in our table, its list price is the highest.

Business Pill · THE JEVONS PARADOX

A one-minute explainer of the Jevons Paradox: why a cheaper resource can lead to more total spending on it. It teaches the general idea only and says nothing about any company in this piece.

What OpenAI Launched

OpenAI’s post describes the problem the product targets: “Applications and agents make many small decisions throughout a workflow: how to classify an input, where to route a request, or which action to take next.” Developers define a question and its allowed answers, and the API returns a probability, a choice or a score.

Since the public beta, OpenAI says it has added “a native GPT‑Live integration, prompt caching for reused context, and inference residency in the US and EU.” Developers can now “ask up to 400 questions per request, twice the previous limit.” On price, the post says: “Decisions API costs $0.10 per million input tokens. Cached input costs $0.01 per million tokens.” It adds: “You only pay for input tokens: there are no output-token charges.”

TechCrunch quotes Nikunj Handa, the product lead for OpenAI’s APIs: “The big unlock with Decisions API has been the speed, shrinking things from one to three seconds to 100, 120, 180 milliseconds.” TechCrunch reports that OpenAI did not build a specially calibrated model; Handa told it that “right now, it’s just plain old Luna that gets the job done.”

How the Prices Compare

Each company lists output tokens as free. Prices are list prices on the dates shown and may have changed since. The cost of one decision depends on how many input tokens it uses, which differs by model and task, so the per-token ratios below are not per-decision costs.

• About 2.4 times Jev. $0.10 against $0.042 per million input tokens is about 2.4 times, by our arithmetic. Microsoft lists the same $0.042 for Decision-1.
• About 2.6 times Cloudflare. Against Clef-flash at $0.038, the ratio is about 2.6 times, by our arithmetic.
• Caching changes the picture. OpenAI’s cached input price of $0.01 per million tokens is below every uncached price in the table. How much of a workload can be cached depends on the application.

Where OpenAI Ranks on Public Tests

TechCrunch reports that public benchmarks “tend to favor D1 Ultra, a new decision model that Liquid released last week, and Base Lab’s Elephant-1.” It adds that the latest Jev model ranks seventh and third on two indices, “with OpenAI’s Decision Model falling eighth and fourteenth, respectively,” and that TypeSafe’s models are “the most widely-queried” on OpenRouter.

TechCrunch also names early customers: “In initial trials with customers like Box, Datadog, Ramp, and Speak, engineers used the API to replace OpenAI’s LLMs with higher throughput.” TechCrunch reports that Speak’s CTO, Andrew Hsu, said the API allowed the company to cut “the time between actions from 1.4 seconds to .13 seconds.”

Connecting the Coverage

• Our piece on TypeSafe’s funding covers the round that valued the Jev maker at $7.5 billion, which TechCrunch reported on 9 October 2026.
• Our piece on OpenAI’s Decisions API pricing covers the same $0.10 per million input tokens when the API was in public beta.
• Our piece on AI labs becoming services firms covers OpenAI’s forward deployed engineers, the team that builds with customers before a product exists.

The Structural Read

As we read it, OpenAI is selling speed and breadth rather than the lowest list price. Its post highlights GPT-Live voice integration, images, residency in the US and EU, and up to 400 questions per request.

As we read it, caching is the lever buyers will test first: OpenAI lists cached input at $0.01 per million tokens, below every uncached price in our table.

As we read it, the category is crowded within weeks: TypeSafe, Microsoft, Cloudflare, Subquadratic, Nace and now OpenAI all list decision models with free output tokens.

Three Implications

FOR DEVELOPERS OpenAI lists $0.10 per million input tokens with cached input at $0.01 and no output charges, so the effective price depends on how much context an application reuses.

FOR TYPESAFE AND OTHER SPECIALISTS TechCrunch reports that the latest Jev model ranks seventh and third on two public indices, against eighth and fourteenth for OpenAI’s Decision Model, and that TypeSafe’s models are the most widely queried on OpenRouter.

FOR BUYERS COMPARING MODELS List prices per million input tokens do not give the cost of a decision, because models use different numbers of tokens for the same task; buyers have to measure that on their own workloads.

The Business Engineer Lens

This story maps onto the Business Engineer framework The State of The Inference Economy.

The framework says: “Inference is a revenue engine: continuous, distributed across millions of applications and agents, incurring marginal cost with every query.”

As we read it, decision models are a direct bet on that marginal cost: each one prices the many small choices an agent makes per task, and OpenAI’s launch adds the largest lab to a price list that already includes Microsoft and Cloudflare. The framework was published in March 2026 and does not discuss the Decisions API or TypeSafe.

What Is Not Established

The price comparison uses list prices per million input tokens. It does not show the cost of a decision, because each model uses a different number of tokens for the same task. Nace has published per-decision costs it measured for its own model and others; we have not seen an independent, common per-decision benchmark. Rival prices are as listed on the dates in the table and may have changed.

The benchmark ranks are as reported by TechCrunch; we have not run the tests. TechCrunch puts TypeSafe at $7.6 billion on 22 October and at $7.5 billion on 9 October; we use the 9 October round figure. The 2.4 and 2.6 times are our arithmetic.

The Bottom Line

OpenAI made its Decisions API generally available on 22 October 2026 at $0.10 per million input tokens, with cached input at $0.01 and no output charges. That is about 2.4 times the $0.042 TypeSafe lists for Jev, by our arithmetic. OpenAI says the API is up to 10x faster than GPT-6 Luna through its Responses API, and TechCrunch reports it ranks eighth and fourteenth on two public decision-model indices.