OpenAI has quietly rolled out an update to its ChatGPT model picker for Pro subscribers, introducing six distinct response tiers. The change gives paying users more granular control over how the chatbot handles their requests. The new options appeared in the interface without a formal announcement, leaving users to discover the shift on their own.
What the new tiers offer
The model picker now shows six tiers instead of the previous handful of choices. Each tier appears to adjust the depth of reasoning, verbosity, or style of the output. Early reports from users on social media suggest the tiers range from concise, quick answers to longer, more detailed breakdowns that take additional processing time. The exact labels and behavior for each tier have not been officially documented by OpenAI, but the company typically updates its help pages shortly after such releases.
Why Pro users get the upgrade first
ChatGPT Pro is OpenAI's $200-per-month subscription tier. It offers priority access, longer context windows, and earlier feature rollouts. This update fits that pattern. The model picker change doesn't affect the free or Plus tiers for now. That's consistent with how OpenAI tests new controls on its highest-paying customers before expanding them more broadly.
How the picker used to work
Before the update, Pro users could choose between GPT-4, GPT-4 Turbo, and a few specialized models. The selection was mostly about which model to use, not how the model should behave. The new tiers move the control closer to the user's specific need — a shift from model choice to response tuning. It's a subtle but meaningful change for power users who rely on varied outputs for different tasks.
What comes next
OpenAI hasn't said when or if the six-tier system will reach non-Pro subscribers. The company also has not published a changelog explaining each tier's intended use. For now, Pro users can experiment with the new options and figure out which tier works best for coding, writing, research, or casual conversation. The lack of official documentation leaves room for community guides to fill the gap — something that has happened before with OpenAI's feature rollouts. Until the company clarifies the specifics, users will have to test the tiers themselves.
