OpenAI added a Dot agent and GPT-6.1 Sol model to Pro

OpenAI's Pro plan gained a new always-on agent called Dot and access to a new GPT-6.1 Sol model, which Plus subscribers can also use. Response times on eligible Pro tiers moved from the previous 'Fast' speed to a new 'Ultrafast' tier, and Codex and ChatGPT work sessions were reframed as running longer rather than being described by a usage multiplier.

Pro's usage limits were also restructured: per OpenAI's pricing page as of Sep 30, 2026, OpenAI replaced the '5x more usage' multiplier framing with a choice of three usage tiers, and refreshed the taglines on all plan cards -- Free, Go, Plus and Pro -- to a consistent 'Best for...' format.

For product leaders, this bundle shows OpenAI leaning harder into agent and model differentiation for its top plan while also changing how usage limits are communicated to subscribers. Moving from a bare '5x more usage' multiplier on OpenAI's pricing page before Sep 30, 2026 to named usage tiers is a packaging change as much as a product change -- it gives OpenAI room to adjust the underlying limits without re-explaining a single number, and it tracks a broader shift from usage-based framing toward tiered, segment-friendly language.

For teams building their own seat- or usage-based pricing, the lesson is the choice of tiers over multipliers: a multiplier invites customers to do the math and compare; named tiers let you reposition limits without a visible cut. Watch whether OpenAI eventually discloses the actual usage caps behind the three tiers, or keeps them as opaque as the multiplier framing did.