Skip to content
AI Radar by Wagner Solution
CS
Menu
important Coding verified update

Decisions API from OpenAI adds voice and image support

independently confirmed updated October 22, 2026

OpenAI has made Decisions API available to all customers and added voice and image support. According to the CTO of Speak, the service reduced the wait between actions from 1.4 to 0.13 seconds. The limit has increased to 400 questions per request.

OpenAI has made the public version of Decisions API available to all customers, with support for voice and images. The service is designed to make quick decisions among specified options. The company has also distributed computing resources geographically to reduce latency across regions.

According to OpenAI, the service uses the ChatGPT Luna model, optimized for faster inference and lower costs. Andrew Hsu, CTO of Speak, said that using the service reduced the wait between actions from 1.4 to 0.13 seconds. This is the result of a specific deployment at a company developing language-learning assistants, rather than a guaranteed latency for all tasks.

The changelog from OpenAI lists general availability with the gpt-6-luna model. The limit has increased from 200 to 400 questions per request compared with the public beta. Shared input can be cached across requests to take advantage of prompt-cache discounts, instructions for individual questions can include images, and structured data can be passed directly through input_json.

What changed

Why it matters

For applications that repeatedly choose among several options, shorter waits could speed up subsequent actions, for example during voice-based tutoring. The result at Speak demonstrates a concrete benefit in one deployment; speed and suitability for other tasks need to be verified using data from those tasks.

What was added since the original report

Verified updates

  1. New verified information

    The public version adds voice support.; OpenAI has distributed computing resources geographically to reduce latency.; According to the CTO of Speak, the wait between actions fell from 1.4 to 0.13 seconds.; The model was optimized for faster inference and lower costs.

    • The public version adds voice support.
    • OpenAI has distributed computing resources geographically to reduce latency.
    • According to the CTO of Speak, the wait between actions fell from 1.4 to 0.13 seconds.
    • The model was optimized for faster inference and lower costs.

Two audiences, two different impacts

What this means

01

For individuals

The higher limit allows developers to combine a larger set of questions into a single request and pass structured data directly through input_json during integration.

What to do Try passing structured data through input_json using a small set of your own questions.
More practical updates →
02

For a business

Companies running interactive assistants have another option for reducing the wait between steps. For shared input used repeatedly, they can also take advantage of prompt-cache discounts; the sources do not substantiate specific savings.

Development
What to decide In a limited pilot, compare latency and decision accuracy with the existing solution on the same tasks.
More business impacts →
ChatGPT Luna Decisions API GPT-6 Luna input_json OpenAI Speak

Check the original

Event sources

independently confirmed · 2 publishers, 1 independent. We count feeds from the same owner only once.

2
OpenAI API Changelog primary source · first detected The [Decisions API](https://developers.openai.com/api/docs/guides/decisions) is now generally available with `gpt-6-luna`. Cache shared input across requests for prompt-cache… TechCrunch AI independent context OpenAI makes its Jev competitor widely available as decision model competition heats up