OpenAI’s GPT-6 Luna-Powered Decisions API Is Now Generally Available for Developers With Native GPT-Live Integration: Price, Capabilities

OpenAI's Decisions API has been available in public beta for a while. Now, the Sam Altman-led tech giant has announced that its Decisions API is now generally available to users. The specialised model interface allows developers to use the tool to build their AI agents and apps, enabling the software to answer recurring questions and make small decisions. Decisions API can use images, text, and other structured JSON inputs from developers. It is powered by OpenAI's GPT-6 Luna AI Model. The company says that Decisions API now ships with native GPT-Live integration, too. Since its public beta release, the company has also added prompt caching for reused context in select markets.

Decisions API Price, Capabilities

In a blog post on Thursday, OpenAI announced that the Decisions API is now generally available in select global markets for developers. It is available in the Decisions Playground. The specialised model interface, powered by the GPT-6 Luna, costs $0.10 (roughly Rs. 10) per million input tokens. Meanwhile, cached input costs $0.01 (about Rs. 1) per million tokens. Decisions API is available for eligible customers with Zero Data Retention.

Developers can use the GPT-6 Luna-powered Decisions API to help their AI agents and apps make small and recurring decisions, like categorising and arranging information, routing requests, and selecting the next action. The AI giant claims that the Decisions API can make “near real-time decisions” using images, text, and structured inputs up to 10x faster than the GPT‑6 Luna AI model via the Responses API.

Moreover, OpenAI claims that developers can pass native JSON inputs and include images in question instructions, while also being able to ask up to 400 questions per request with the Decisions API. This limit has been increased from up to 200 questions per request since the specialised model was launched in public beta. Along with increasing the limit, the company has added a native GPT‑Live integration, prompt caching for reused context, and inference residency in the US and EU.

A developer has to provide the questions, a set of relevant answers, and context to the Decisions API, which can choose the most appropriate answer to the query based on the context. Hence, it provides an actionable result to the app or AI agent. With the native GPT-Live integration, the Decisions API now helps apps, voice agents, and other AI-powered customer support agents decide whether there is a need to ask clarifying questions, redirect requests, or escalate any particular issue.