Deep Review
Published Updated
What is Deep Review?
A more thorough AI review of a queued prompt, typically run on Opus high-effort or another premium model. Used when stakes warrant extra scrutiny.
Deep Review uses a more capable (and expensive) model to evaluate a draft prompt — typically Opus high effort. Configured per tier; high-complexity tier defaults to manual reminder, where the user can choose between Auto Review (Flash) and Deep Review (Opus) depending on the situation.
How it relates to Orbit
Orbit’s Deep Review adds a stronger model review to a queued prompt before you approve it. See Queue & Approval for details.
When you would use it
You run Deep Review when a prompt's blast radius is large enough that a thorough review pays for itself — typically refactors, architectural changes, or anything affecting production.
Common operations
- Reviewing high-stakes prompts with deeper analysis
- BYOK premium-model review for important work
Related terms
Where this is taught
No learning path uses this term yet. Browse the Learning Atlas for guided sequences through related ideas.