OpenAI used DevDay in San Francisco to announce Dots, an always-on AI assistant, alongside a faster Codex tier and a new $500 monthly plan. The announcements point to a broader push: AI that works in the background, collaborates across shared spaces and charges a premium when speed matters.
OpenAI Watch analysis
What happened
Mashable reports that Dots can work across existing context and connected applications, learning from user feedback as people use them. The assistant runs in the background through a cloud computer, OpenAI says, and is rolling out to eligible Pro and Business Premium users.
OpenAI also announced GPT-6.1 Sol, which Mashable says is priced at $2 per million input tokens and $10 per million output tokens. Its new Pro 500 plan costs $500 a month and includes Ultrafast, a speed tier for Codex and Work that can generate up to 300 tokens per second. The company says Ultrafast is also coming to the API, at up to six times standard speed. ChatGPT Space gives colleagues a shared workspace where their Dots assistants can access common context.
Why it matters
This is more than another model name on a crowded launch-day list. Dots is pitched as an assistant that stays active across apps, while Spaces gives it shared workplace context. That raises the practical stakes around what an agent can access and do, not just how clever its replies sound.
The pricing also makes the trade-off unusually plain. Sol is positioned as a cheaper model, while Pro 500 sells speed at a premium. At $500 a month, “faster” is a product decision with a very visible invoice attached.
Our read
The most consequential idea here is the always-on assistant, not the mascot. If Dots can reliably use connected apps and shared context, it could make AI more useful in daily work. But the announcement’s promise is not the same as proof of dependable performance, and users will want clear controls over what a persistent assistant can access.
For now, treat the plan and rollout details as useful signals of where OpenAI is taking its products. The real test is whether Dots saves people effort without creating a new job called supervising Dots.
What to watch
- How quickly Dots reaches eligible users, and which connected applications it supports.
- What permissions and user controls are available for its background work.
- Whether GPT-6.1 Sol delivers useful savings on real workloads, not just lower listed prices.
- Whether the $500 Pro 500 plan and Ultrafast tier attract enough demand to become more than a premium niche.
Discussion spark: Would you trust an always-on assistant with connected apps and shared work context if its permissions were clear, or should background agents require approval for every consequential action?
Sources and evidence
- Biggest news from OpenAI Dev Day: Dots, GPT-6.1 Sol, new plans – Mashable (7 October 2026, 10:20 UTC)
OpenAI Watch is independently operated by WittyWires. It is not affiliated with, endorsed by, or operated by OpenAI.