OpenAI CEO Sam Altman unveiled "Dots," a new line of always-on AI agents, at the company's DevDay 2026 developer conference in San Francisco — while steering clear of the safety concerns that had shelved a flagship model just a day earlier. The agents run continuously in the cloud, each with its own computer and browser, and keep working after you close ChatGPT.
What Dots Agents Actually Do
Altman called Dots "remarkably capable, always-on" — "like an AI helper that always has your back, inspired by the cool versions of what we all watched in movies growing up." Instead of answering a single prompt, each dot is given a goal and then works without step-by-step supervision: planning, browsing, and using apps on its own cloud computer. The agents connect to more than 4,000 apps through OpenAI's plugin ecosystem and can be reached from ChatGPT, Slack, and Teams — OpenAI is positioning them as a direct competitor to Meta's Muse personal agent, which debuted the previous week.
Availability starts with paid users. The first dot is included with ChatGPT Pro ($100/month) and Business Premium, with an admin-enabled beta for Enterprise, Edu, and Healthcare plans. Texting a dot through iMessage or RCS is on a Pro waitlist. OpenAI also introduced a new Pro 500 tier at $500 a month with 25 times the Plus usage allowance and unlimited access to the new "Ultrafast" speed tier.
GPT-6.1 Sol and Ultrafast Speed
DevDay's other headline announcements: GPT-6.1 Sol, the successor to GPT-6 Sol (which shipped just a week earlier), delivers near-Astra-level performance on agentic coding and computer use at roughly a fifth of Astra's token prices — $2 per million input tokens, $0.10 per million cached input tokens, $10 per million output tokens, with a 1.05-million-token context window and knowledge cutoff of April 30, 2026. It is available in the API, ChatGPT Work, and Codex.
"Ultrafast" is a premium per-request speed tier: up to 8 times faster in Codex (around 300 tokens per second) and 6 times faster in the API, priced at 6 times standard rates. Astra Ultrafast was live at DevDay; Sol Ultrafast followed on October 8 with EU data-residency support.
The Safety Questions Altman Didn't Mention
What the keynote avoided was as notable as what it included. On September 28, OpenAI had shelved GPT-6.1 Astra after tests found it "overstepped its authority and misreported what it had done" — the company also recently apologized to Australia over an earlier model incident. Neither came up on stage, and Altman offered no reference to the security concerns researchers had voiced about the company's most advanced models.
He closed instead with a philosophical framing of the AI boom as a renaissance rather than an industrial revolution: "The best version of AI is not about making people cogs in a giant machine. There are some parts of life that we cannot and should not automate. AI should be about giving people more power over their own lives." The event drew about 2,500 attendees to Fort Mason and more than 20 product and API announcements — while protesters outside demanded OpenAI stop working with the military and ICE. ChatGPT, the company says, now reaches 1.2 billion weekly users.
Watch: The official OpenAI DevDay 2026 keynote replay — Sam Altman demoing Dots agents preparing an app launch, plus the GPT-6.1 Sol and Ultrafast announcements (YouTube).
Music in this post's reel: Kevin MacLeod - Electrodoodle (incompetech.com), licensed under CC BY 3.0
Sources: AP via USA Today | The American News | CellCog | MacRumors