Fast requests being used with Free model - Gemini 2.5 flash preview 5-20

Describe the Bug

I have noticed today that I have started to incur fast requests when using Gemini 2.5 flash preview 5 20 which is the free model. When I mouse over the model in cursor it shows as 0xrequests for the cost. Its using around 0.75xrequest tokens per request.
When I view my cursor dashboard it doesn’t show any premium usage and confirms its registering the model used as Gemini 2.5 flash preview 5 20. I’ve made sure I’m not using any premium models (deselected them all in cursor settings).

When I select my past chats in cursor and mouse over the most recent chat it shows that I’m using fast responses. I also checked my previous chats in cursor from the last few weeks which ALL confirm that no fast requests were made then when using the free Gemini model

has anyone else seen this, I’ve only came across this today and contacted cursor to query it. I’m currently waiting for their response and they recommended I pop over here to see if anyone else has had similar issues.

Steps to Reproduce

Attempt to use Gemini flash preview. observe your fast responses usage in chat history (mouse over pop-up)

Expected Behavior

No fast responses should be used when using the 0xrequest cost model

Operating System

Windows 10/11

Current Cursor Version (Menu → About Cursor → Copy)

version 1.1.3
vscode version 1.96.2

Does this stop you from using Cursor

No - Cursor works, but with this issue

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.