ChatGPT Free and Go tiers now use the fastest models by default

OpenAI has quietly flipped a switch that changes how millions of people interact with ChatGPT. If you are currently rocking a free account or that budget-friendly “Go” subscription, you are no longer going to get the premium treatment where the app automatically picks the smartest brain for your question. Instead, you are getting defaulted straight to GPT-5.2 Instant.

You’re going to have to ask for the smart model nicely now

The sole official announcement for the update was available in the release notes for ChatGPT. It went as follows:

“Update for free and Go users: We’re removing automatic model switching for reasoning in ChatGPT. Previously, some questions were automatically routed to the Thinking model when ChatGPT determined it might help. To maximise choice, free users will now use GPT-5.2 Instant by default, and can still choose to use reasoning anytime by selecting Thinking from the tools menu in the message composer.”

Basically, the company rolled back its “model router” for most of us. This was the automated system introduced back in August with GPT-5. It decided if your prompt needed a quick answer or a deep dive. Now, the app will just serve up the Instant model. It is snappy and cheap to run, but it isn’t exactly Einstein.

Of course, you can still access the “Thinking” model if you really need it by hitting the + icon in the chatbox. But the catch is that you have to now manually select itevery single time. It is a bit of a friction point, and perhaps OpenAI thinks most people won’t bother.

ALSO READ: ChatGPT 5.2 lands with big upgrades for professional work

Beyond that, the update likely also has something to do with saving cash.

Reasoning models burn through massive amounts of compute power, since they run queries upon queries to check and double check their own responses. Serving them to free users could have been costing the company a fortune.

That’s perhaps even more true with the all-new ChatGPT Images models and all of the processing they likely require.