Platform

AI Dashboard

Health across all 2 workspaces.

Risk assessment (.md)

Model calls: 120; 2 workspaces

Eval gate ≥80%: 60%; latest of 3 runs

Over cost ceiling: 6; 5.0% of calls

Fallbacks: 0; 0.0% of calls

Total cost: $1.27; 120 calls

29 Jun 2026 to 14 Oct 2026

By model

Claude Sonnet 4.651 calls · $0.51
GPT-4.119 calls · $0.06
GPT-4o mini19 calls · <$0.01
Claude Haiku 4.519 calls · $0.05
Claude Opus 4.812 calls · $0.65

Eval pass rate

Gate 80% · latest 60%