Codex scans repos and decides faster

The Decoder reports that OpenAI expanded Codex and its API at DevDay 2026 in San Francisco. The security piece is Codex Security Cloud. It is aimed at security teams. It scans entire GitHub repositories on demand or on a schedule, and it keeps checking new commits. Codex investigates what it finds, filters out duplicates, and prepares fixes in the cloud, including while a laptop is closed. In July, OpenAI released the Codex Security CLI as an open-source tool for the command line and for CI. The cloud version is available to Pro, Business, Enterprise, and Edu, on desktop and web. Those users also get the Daybreak Blue models without a separate application. Daybreak Blue is the defensive tier of OpenAI’s cybersecurity program, for tasks such as malware analysis. It previously required identity verification, among other things.

The Decisions API uses a version of GPT-6 Luna, the lower-cost model. Developers define a question with a limited set of predefined answers, pass context as text or images, and get an answer back. Uses include classifying content, routing requests, and choosing an agent’s next step. It launches as a limited preview, with broad availability expected in the coming days. OpenAI says the Decisions API makes decisions ten times faster than GPT-6 Luna does through the regular API. That is the company’s claim, shown with a figure, not a measurement repeated here. The same piece says the Agents API adds Computer Use, so agents can operate software themselves. That capability has been in Codex and ChatGPT on macOS since April and on Windows 11 since May. The API has been in public beta since September 11.

Ultrafast is a paid speed tier. OpenAI says it speeds token generation in Codex by up to eight times, to 300 tokens per second, and by up to six times in the API. API pricing is six times the standard rate. For GPT-6 Astra that is $60 per million input tokens and $300 per million output tokens. The company also offers Astra’s Fast mode at double the price for 2.5 times the speed. GPT-6 Astra Ultrafast is available now in the API, and in ChatGPT Work and Codex for Enterprise customers and for subscribers to the new Pro 500 plan. Pro 500 costs $500 a month. It is supposed to offer the highest usage allowance of any subscription, though OpenAI does not say how high. The existing $200-a-month Pro plan will include only ten times Plus usage in ChatGPT Work and Codex, instead of twenty times. GPT-6 Pro messages in chat are cut from 200 a week to 100. Existing customers keep their current limits for a transition period and receive a one-time credit.

The same set of updates includes reusable cloud development environments. Codex in the cloud is available on Plus, Pro, Business, Enterprise, Edu, and Healthcare. Teams can share one configuration with approved settings and permissions. The command line can also start and steer tasks by voice, and the new CLI is on all plans. The desktop app adds a code-review view across projects, also on all plans. Those are accompanying capabilities. A reader watching price should start with Ultrafast at six times the rate and with the change in Pro allowances. A reader watching repositories should start with Security Cloud scanning GitHub and preparing fixes in the cloud.