if I select a model when I create a project in the cloud, I discover later that it’s using different models than my preferred. Oftentimes, more expensive models are being used.
For AI issues: which model did you use?
Model name (e.g., Sonnet 4, Tab…)
For AI issues: add Request ID with privacy disabled
Request ID: f9a7046a-279b-47e5-ab48-6e8dc12daba1
For Background Agent issues, also post the ID: bc-…
Additional Information
Add any other context about the problem here.
Does this stop you from using Cursor?
Yes - Cursor is unusable
Sometimes - I can sometimes use Cursor
No - Cursor works, but with this issue
The more details you provide, the easier it is for us to reproduce and fix the issue. Thanks!
I looked at your recent Projects activity. The project itself ran on the model you selected (Grok 4.6 High). What you’re seeing are the worker agents the project coordinator launches to do parts of the task in parallel. Workers are supposed to run on the project’s model unless you ask for a different one, but right now the coordinator can still choose another model on its own, and over the last two days it started several workers on faster, more expensive variants (GPT 5.6 Sol Fast, GPT 5.6 Terra Fast, Grok 4.6 Fast). This is a known issue we’re already tracking, and I’ve added your report to it.
Two things that help today:
Add a line to the project’s prompt or rules such as: “Start every worker agent on the same model as this project. Do not pick a different or faster model unless I explicitly ask for one.” This is guidance rather than a hard setting, but it makes a real difference.
If you do want a specific model for the workers, name it explicitly in that same line (for example “use Grok 4.6 High, non-fast, for all workers”). Note that disabling models under Settings > Models only filters the picker in the desktop app and does not restrict cloud workers.