I find it really hard (every couple of days) to write a post to ask basic questions because of the lack of precision / clarity from the Cursor team. For example, in the Changelog changes, we don’t know which changes are on 0.45.6, and which are on 0.45.7… The Cursor team only cares about the latest version, like if other users who have previous versions don’t matter (and they need to know which features/changes are included in their own version, which is not the latest)…
Now that o3-mini is out, Cursor supports it, but we don’t know which version, in other words, o3-mini has a parameter “reasoning_effort” which affects the quality / accuracy / speed, etc… and we have no idea which one is used (cf. https://platform.openai.com/docs/guides/reasoning#reasoning-effort)
I really hope the Cursor team puts more effort into their communication and clarity; this would prevent many posts where users legitimately ask questions from time to time
hey! thanks for this feedback. we definitely have room to improve on how we communicate the changes that go out in patch version, you’re totally right. to give a little background, we’re shipping a lot of features and fixes multiple times per day which makes it very time consuming to keep track and making sure this is accurate and up to date. most focus go on improving the product!
with that said, it’s definitely something on our radar and expect us to get better at this in the future. we really want to be clear, set the right expectations and communicate more. again, thank you for this feedback!
to answer your question – we’ve just updated to high reasoning for o3. let us know what you think!
Thanks a lot for considering this and for your feedback Really appreciate it!
It’s good news that you have now updated to o3-mini (high)! Will def try it!
Really appreciate it!