OpenAI announces Codex cloud environments, Decisions API and Ultrafast tier at DevDay 2026

Sep 29, 2026

At its DevDay 2026 conference on September 29, 2026, OpenAI unveils a major shift toward agentic workflows, launching always-on AI agents called Dots and persistent Codex cloud environments. The lab also introduces a premium Ultrafast speed tier alongside a $500 monthly Pro plan and the new Decisions API. Amid these product rollouts, executives address safety concerns that shelved their flagship GPT-6.1 Astra model, while downplaying immediate IPO timelines despite reports of a new $30 billion funding round.

DevDay 2026 product announcements

▪OpenAI announced more than 20 new products and features at its annual DevDay developer conference in San Francisco on September 29, 2026.
▪OpenAI introduced GPT-6.1 Sol on September 29, 2026, a cheaper workhorse model designed for professional work, computer use, and agentic coding.
▪OpenAI previewed OpenAI Private Intelligence on September 29, 2026, which features zero data retention and private safety processing to protect sensitive user data.
▪OpenAI shelved its plans to launch the GPT-6.1 Astra model on September 28, 2026, because the model did not meet the company's safety standards.
▪OpenAI introduced ChatGPT Space on September 29, 2026, a collaborative workspace allowing teammates and AI agents to work together on shared project knowledge.

Codex cloud environments

▪OpenAI's updated Codex Command Line Interface, released on September 29, 2026, allows developers to start and direct tasks using voice commands
▪OpenAI launched Codex Security Cloud on September 29, 2026, a toolset that scans GitHub repositories for vulnerabilities and prepares fixes in the cloud.
▪OpenAI updated its software engineering agent Codex on September 29, 2026, introducing persistent, reusable cloud development environments accessible from any device.

Ultrafast tier pricing

▪OpenAI's Ultrafast speed tier is priced at six times the standard rate of its API, costing $60 per million input tokens and $300 per million output tokens for GPT-6 Astra
▪OpenAI introduced a premium speed tier called Ultrafast on September 29, 2026, which generates tokens up to eight times faster in Codex and six times faster in OpenAI's API
▪OpenAI launched a Pro 500 subscription plan for $500 per month on September 29, 2026, which bundles the Ultrafast speed tier and offers the highest usage allowance of any OpenAI plan

Dots always-on AI agents

▪OpenAI launched Dots on September 29, 2026, which are always-on AI agents powered by GPT-6 Astra that run on their own cloud computers.
▪OpenAI rolled out Dots, its always-on AI agents, on September 29, 2026, initially for ChatGPT Pro, Business Premium, and Enterprise subscribers in eligible markets
▪OpenAI's Dots agents can connect to more than 4,000 applications, learn from user feedback, and are accessible within ChatGPT, Slack, and Microsoft Teams

OpenAI IPO timing strategy

▪OpenAI is engaging in early-stage discussions with investors to raise a potential funding round of approximately $30 billion, according to reports on September 29, 2026.
▪OpenAI Chief Executive Officer Sam Altman stated on September 29, 2026, that the company does not have a specific timeline for an initial public offering.
▪OpenAI Chief Financial Officer Sarah Friar stated on September 29, 2026, that the company wants to pursue an initial public offering when the time is right for the business.

Decisions API launch

▪OpenAI's Decisions API uses a version of the GPT-6 Luna model to answer user-defined questions with predefined, finite answers
▪OpenAI stated that its Decisions API processes decisions ten times faster than the GPT-6 Luna model does through OpenAI's standard API
▪OpenAI launched the Decisions API in limited preview on September 29, 2026, designed for real-time decision-making and classification tasks.

Debatable claims

▪OpenAI's high-priced premium tiers create an unfair digital divide in technology access
▪OpenAI's rapid model release cycle compromises its commitment to AI safety