I am wondering if there is a way to configure reasoning effort for AWS Bedrock models (like Claude Opus 4.7)? Now our team has to switch to Claude Code for tasks needing xhigh effort. Same question for GPT-5.5 which is going to be available on AWS soon.
Great question! Unfortunately, it’s not currently possible to set effort levels for Bedrock models in Cursor. I’ve moved your post to Feature Requests so the team can track this suggestion.