OpenAI opens Decisions API to all customers with voice and visual support
By Automate Basics, written with AI from TechCrunch's original
OpenAI has made its Decisions API available to all customers and added voice and visual support, TechCrunch reports. The model is designed for workflows that need to choose quickly among supplied options and return a confidence rating.
What changed
OpenAI has made its Decisions API available to all customers after an earlier announcement at its developer event, according to TechCrunch. The public release adds voice and visual capabilities. The model is meant to make a quick choice from a set of options and provide a confidence rating, rather than generate a long answer.
OpenAI's API product lead told TechCrunch that the company adapted its existing Luna model to respond faster and at lower cost. OpenAI also placed computing resources in different regions to reduce delays. For a workplace team, the distinction is practical: a decision model may fit a repeated choice in an automated process better than asking a general-purpose chatbot to reason through the same options each time.
Who can use OpenAI's Decisions API?
OpenAI's Decisions API is available to all customers, TechCrunch reports. The article does not identify separate eligible plans, give a price or spell out whether voice and visual use have different charges. Teams should check their own OpenAI account terms before estimating the cost of a workflow.
TechCrunch describes the product as an API, which means it is aimed at people building or maintaining software connections, not presented as a new feature to turn on in the ChatGPT app. A nontechnical team might therefore work with whoever manages its automations to assess whether a recurring choice is worth moving to the API. Availability alone does not show whether it will be cheaper for a particular task.
How do I try it at work?
A sensible first trial for OpenAI's Decisions API is a repeated, low-risk choice with a small set of allowed outcomes, such as assigning incoming requests to existing categories. Have the person responsible for the automation compare its selections with decisions your team has already checked. Review the confidence ratings too, especially when requests are unclear or fall between categories.
TechCrunch reports faster responses in initial customer trials, but those results do not establish how the API will perform in every workflow. Check whether it chooses the right option, handles ambiguous inputs appropriately and saves enough time to justify the work of connecting it. Keep a person involved where a wrong choice could affect a customer or an important business decision.
Frequently asked questions
What does OpenAI's Decisions API do?
OpenAI's Decisions API is designed to choose among supplied options and return a confidence rating, according to TechCrunch. It is intended for quick, repeated decisions in software workflows rather than long written responses. OpenAI has also added voice and visual capabilities to the version now available to all customers.
Is the Decisions API included in my OpenAI plan?
TechCrunch reports that OpenAI has made the Decisions API available to all customers, but does not specify plan tiers or provide pricing. The report describes an API for software connections, not a feature inside the ChatGPT app. Check your account terms before planning a rollout or budget.
How should a team judge whether the Decisions API is useful?
Start with a recurring choice that has clear options and results your team can check. Compare the API's selections and confidence ratings with known decisions, including ambiguous cases. TechCrunch reports faster responses in initial customer trials, but those results do not establish accuracy or cost for your workflow.
For you
Learn to do this with AI, free
Short courses for people who aren't engineers: ChatGPT, Claude, prompting and automation. Every lesson is free, with no card.
Start a free courseFor companies
Make an AI or automation tool?
List it in the directory for $9.99, refunded in full if it isn't approved, or claim its page if it is already here.
Related reading
1Password introduces AI agent logins without sharing passwords1Password has introduced a way for AI agents to sign in to a user's personal accounts without receiving the password. The user approves each login request, then 1Password supplies the credentials on the agent's behalf.
Canva launches Craft image model for planned design generationCanva has launched Craft, an image model that works out a design before producing the finished image. It is available to Canva users worldwide and is aimed at tasks such as social posts and marketing graphics.
Google Vault adds filtered exports for Gemini in Workspace chatsGoogle says Vault administrators can now filter and export Gemini in Workspace conversations by user or organizational unit and date range. The change gives teams handling legal or compliance requests a way to collect eligible AI conversations across Workspace apps.
Written with AI from TechCrunch's original and published after automatic checks: every figure here appears in the original, and no sentence is copied from it. The picture is AI-generated. The original is the authority.
Get the next one by email
One email, most weeks: what changed in AI for people who use it at work, and what to do about it.
Source: OpenAI makes its Jev competitor widely available as decision model competition heats up, TechCrunch, 22 Oct 2026.