As in openai document, we need “reasoning_effot” parameter for usage of “o3-mini-high” like below.
Please add model with this parameter or, just add o3-mini-low/medium/high please.
const completion = await openai.chat.completions.create({
model: "o3-mini",
reasoning_effort: "medium",
messages: [
{
role: "user",
content: prompt
}
],
store: true,
});
14 Likes
Was just about to make a post on this. I am curious as to what it is currently set as in cursor.
2 Likes
I am using o3-mini via my API key. How do I pass “high” effort as a parameter. Default is medium. But I would like to control this parameter, if possible.
2 Likes
we currently don’t have a way to control this with custom api keys for now. however, we’ve updated o3-mini
to use high
reasoning. let us know what you think!
2 Likes
Cool! Thanks! I will try to use for my project.
@ericzakariasson I appreciate the acknowledgement that Cursor is using high reasoning. That makes all the difference in whether I choose o3-mini or DeepSeek-R1.
I want to point out that it took me over 30 minutes of searching to come across this forum and this answer. And even still, a comment on a random forum post isn’t really what I’d consider canonical information.
Even if it’s not possible for users to choose the reasoning level, some indication in the UI that it is indeed set to “high” would be incredibly useful when evaluating options as an end user.
Basically, the information about this should be surfaced in a more direct and easy to find way.