Loading…
Loading…
Enforce hard dollar limits, track hourly token burn rates, and prevent runaway LLM inference billing spikes.
| Budget Policy | Scope Target | Spend / Monthly Cap | Hourly Burn Rate | Est. Days to Breach | Enforcement Action | Status | Audit |
|---|---|---|---|---|---|---|---|
GPT-4o Production Monthly Cap bgt-gpt4o-global | model: gpt-4o | $1240.50 / $1500.00 82.7% allocated | $2.14 / hr | 5.2 days | throttle rate | warning | Inspect |
Copilot Agent Autonomous Workspace bgt-copilot-service | service: copilot-agent | $340.20 / $800.00 42.5% allocated | $0.65 / hr | Within limits | hard kill switch | healthy | Inspect |
Claude 3.5 Sonnet Financial Extraction bgt-claude35-rag | model: claude-3-5-sonnet | $2580.00 / $2500.00 103.2% allocated | $4.82 / hr | 0 days | hard kill switch | breached | Inspect |