AI & Agents

OpenAI DevDay 2026: persistent “Dots” agents, a $500 Pro plan and an Ultrafast speed tier

Developing

Details come from release-note summaries; the speed claims are OpenAI's.

The audience at a developer conference session
Photo: Game Developers Conference 2010 (4439258202) by Official GDC, CC BY 2.0, via Wikimedia Commons

In 60 seconds

  • OpenAI's DevDay 2026, on September 29, introduced Dots, autonomous agents with persistent memory, plus synced Work tasks across devices and ChatGPT Space for collaboration.
  • A Pro 500 plan, priced at $500 a month, gives premium access to the Ultrafast speed tier.
  • OpenAI says Ultrafast offers up to 8x faster token generation in Codex and up to 6x in the API.
  • Library connectors for Box, Dropbox and SharePoint and Deep Research in ChatGPT Work and Codex were also listed.

Impact Layer

What changed
OpenAI added long-lived agents and a higher-priced tier aimed at heavy work users.
Why it matters
Agents that remember context and run continuously will do more of the repetitive marketing and operations work, and teams will need rules for what they may do.
Who is affected
Developers, marketing operations teams and enterprise buyers.
Martech stack area
Marketing Automation, AI
Stack impact
Decide which tasks a persistent agent may do unattended, and who reviews its work.
Business impact
Moderate Described from a release-notes summary and a search summary of the event; OpenAI's DevDay page blocked access.
Related technologies
Related vendors
What to watch
Availability of Dots to Enterprise customers, data retention for agent memory, and pricing beyond Pro 500.

OpenAI’s DevDay 2026, held on September 29, brought new agent features and plans, as listed in OpenAI’s release notes and summarised by Releasebot.

The items

  • Dots. Autonomous agents with persistent memory.
  • Work tasks synced across devices, and ChatGPT Space for collaboration.
  • Pro 500 plan. A $500 a month plan with premium access to the Ultrafast speed tier, which OpenAI says gives up to 8x faster token generation in Codex and up to 6x in the API.
  • Astra Ultrafast. Faster generation through WebSocket mode in the Responses API, available to Enterprise and Pro 500 subscribers.
  • Library connectors for Box, Dropbox and SharePoint, and Deep Research in ChatGPT Work and Codex.

Our read

The significant item for marketing and operations teams is Dots: an agent with memory that keeps working. That moves the question from “what can it answer” to “what may it do unattended.” Write that down before piloting: allowed actions, an approver for outside-facing output and a way to review what the agent remembers. The speed figures are OpenAI’s, and the tier is aimed at heavy users. We relied on release-note and search summaries because OpenAI’s event page blocked access.

Sources

  1. Releasebot: OpenAI release notes, September to October 2026 (opens in a new tab) News report

    Source of the DevDay items, Pro 500, Ultrafast figures and connectors. A third-party aggregator.

  2. OpenAI: DevDay 2026 recap (opens in a new tab) Primary source ·

    The event recap. The page blocked direct access, so it was not read; the date and the $500 price come from search summaries.

Get MarTech Roundup in your inbox

What changed in marketing technology, why it matters, and what to do next.

We will email you to confirm. You can unsubscribe at any time.

Related stories