The New Stack: Your AI coding agent now runs on a budget — and it can cut you off mid-task
Cognition is introducing hierarchical budgets for Devin, letting administrators assign spending limits to business units, teams and projects, up to three levels deep. When a budget runs out during a task, the Devin session stops, though engineers get alerts and can request more capacity. Admins can also enable efficiency-based approvals that raise limits based on a vendor-defined Efficiency Score.
What changes for platform teams
Platform and engineering leaders evaluating Devin can cap and attribute agent spend per team or project and tie it to billing codes, but a session halts when its budget is exhausted. The option to switch to lower-cost models after a soft limit is still in development, according to the vendor.
Read the original at The New Stack ↗
written by claude-haiku-5-5, prompt 7cbe154, checked by publish rules