OpenAI DevDay 2026: Dots Are Always-On Agents That Work While You Sleep

OpenAI held its annual DevDay conference in San Francisco on September 29, 2026, and this year's event felt less like a product launch and more like a new company identity: OpenAI is now, unambiguously, an agent platform. The company announced over 20 new features, but three define where the company is heading.

Dots: AI Agents That Never Clock Out

The headline feature is Dots—always-on AI agents that live inside ChatGPT. Each dot gets its own cloud computer and browser, runs on the GPT-6 Astra model, and works toward goals you assign to it around the clock. Think of a dot as a persistent, autonomous assistant that picks up tasks while you're asleep and reports back when it's done. According to 9to5Mac, dots are rolling out now to Pro and Business Premium subscribers, with Enterprise, Edu, and Healthcare availability following after an admin opt-in.

GPT-6.1 Sol: Astra-Level Intelligence at 20% of the Cost

OpenAI also introduced GPT-6.1 Sol, an upgrade to its GPT-6 Sol model that brings it "close to Astra-level intelligence" at just $2 per million input tokens and $10 per million output tokens—roughly 80% cheaper than GPT-6 Astra. The company says GPT-6.1 Sol is particularly stronger on agentic coding, computer-use tasks, and professional workflows. In a notable move, OpenAI simultaneously pulled its planned GPT-6.1 Astra launch because the model did not meet the company's safety standards—a rare public admission of a shelved release.

ChatGPT Space and Pages

For teams, OpenAI introduced ChatGPT Space: a shared workspace where human teammates and their dots can collaborate. Inside Space, a new document format called Pages combines charts, images, checklists, and dashboards in a single editor that humans and agents can edit together. It's a direct answer to emerging competition from tools like Notion AI and Microsoft 365 Copilot agents.

Ultrafast Mode

Rounding out the announcements is Ultrafast, a new premium speed tier for the API and Codex: up to eight times faster in Codex and six times faster in the API. It targets time-sensitive agentic pipelines where latency compounds quickly across multi-step tasks.

With dots, GPT-6.1 Sol, Space, and Pages now in place, OpenAI's message from DevDay 2026 is clear: the standalone chatbot era is over. The next phase is agents you delegate to, not prompts you type.