Which model is the fast request consumed here? This was just discovered after updating 0.45.1, not quite sure~
Features → Large context
When enabled, Composer / Agent use longer context windows. This may use more fast requests.
Likely Sonnet.
The option is in 0.45.0
I haven’t updated as some people are getting rollbacks to 0.44
2 Likes
Yes, the official website has rolled back, but my Cursor automatically updated, haha, so I’m quite curious about what this feature is for~
I had 0.45.1 but was forced back to 0.44
Hey, once v0.45.1 is fully released, you’ll find some more information about this in our changelog!
Actually, I want to know how this Large Context works. Because there is no place that allows me to see whether my context is a Large Context, and how long the context actually is. I have already upgraded to 0.45.2 using other methods. It is completely unnoticeable. Users have the right to know whether this is actually working and whether it is working in the way we understand.
Otherwise, everything is still a game of rock-paper-scissors.
1 Like
We do have some UI coming soon to help show this better, but we wanted to get some backend improvements out first!
More details on the current implementation will be available in the changelog, once more users have the update!
Hey, this is a known bug we are currently trying to fix unfortunately, your best bet is to start a new Composer session to allow the AI to see the file!
Thank you for the follow-up. I hope you guys iron it down soon.