OpenAI DevDay 2026: Dots Agents Ship, GPT-6.1 Sol Undercuts Astra, $500 Pro Tier — and an IPO-Until-Safe Pledge
September 29, 2026, Fort Mason, San Francisco. OpenAI claims "more than 20 major announcements." What actually shipped, per OpenAI's own recap and The Verge:
The headline items
- Dots — always-on personal AI agents, the product that answers Meta's Muse (launched earlier this month) and Google's Gemini Spark (May). Powered by GPT-6 Astra, each Dot gets "its own cloud computer to access a web browser and more than 4,000 supported apps," connects to Slack and Microsoft Teams carrying context across sessions, and is reachable by voice call. OpenAI's pitch: "The magic of dots is when they bring you work done the way you would do it, sometimes before you even think to ask." Available on Pro and Business Premium (Enterprise/Edu/Healthcare get an admin-gated beta, off by default); CFO Sarah Friar told CNBC the vision is to bring it "to our whole consumer base as well." Notably, Dots ship with a "read-only" mode that prevents agents from controlling a user's browser or computer when he or she is not present — a safety control clearly designed for the post-rogue-agent moment.
- GPT-6.1 Sol — the 6.1 generation ships one day after the frontier version, GPT-6.1 Astra, was scrapped over safety findings (see The Pacing Debate Ships Product: OpenAI Kills GPT-6.1, Nvidia Ships Agent Containment, the Market Starts Pricing the Slowdown). OpenAI's recap: it "delivers near-Astra intelligence to everyone at a fifth of its standard input and output token prices." The pacing debate, expressed as a price sheet.
- Pro 500 — the new $500/month ChatGPT tier (25× the Plus allowance) including Ultrafast (up to 8× faster token generation, 300 tokens/second, in Codex; 6× in the API). The $200 Pro tier reopened with new usage limits.
- 1.2 billion weekly ChatGPT users — up from the 1 billion milestone crossed in July.
- Daybreak surfaced: the security deployment product from the pre-keynote rumor list now lives inside Codex Security Cloud, which "includes access to models offered through Daybreak Blue without a separate Daybreak application," and scans GitHub repos on demand with cloud-side remediation.
- Also shipped: ChatGPT Space (collaborative team spaces — the rumored "Spaces"), Bedrock Managed Agents powered by OpenAI (OpenAI agents running natively inside AWS — a significant distribution deal with Amazon), an OpenAI Marketplace with 32 enterprise partners (Adobe, Salesforce, Harvey, CrowdStrike…), Agents API with computer use, and a Decisions API built on a model called Luna.
Rumor scorecard vs. the pre-keynote watch: Dots ≈ the rumored "Aeon"/"o" always-on assistant; ChatGPT Space ≈ "Spaces"; Pro 500 ≈ the "$500 ChatGPT Pro Max" plan; Daybreak shipped (via Codex Security Cloud). No GPT-6 Cyber preview appears in the recap.
The IPO bombshell
Altman, in a Q&A after the keynote (The Verge): OpenAI won't go public until it can make confident safety claims about its models — no timeline:
"We intend to continue with AI progress … but as the models have had this surge forward in capability, and we see more of that ahead of us, we have got to be able to make confident safety claims."
He added that waiting too long for an IPO would be "bad for the world." This hardens September's "ill-advised moment" framing into a stated condition — safety claims as the gating criterion for a listing.
Color
DevDay opened with protests sponsored by more than a dozen organizations (SEIU, Tech Workers Coalition, SF Labor Council, Bay Resistance): "Sam Altman, get off it, put people over profit," with signs reading "No killer robots for ICE," "Where is your moral codex?" and "No climate destruction."