Demo mode— seeded data, no backend. Point FORGE_API_URL at a live agent layer to see real workflows.

LLM costs

Every model call flows through one gateway that logs tokens, latency, and dollars — and hard-stops at the daily budget.

Today's budget

$3.61 of $8.00

Calls are refused once the budget is reached (hard stop, resets at UTC midnight).

Daily spend (last 14 days)

2026-06-262026-07-09

By service

ServiceCallsCost
workflow-generator214$1.9123
diagnostician96$0.8842
workflow-validator402$0.4011
run-monitor611$0.2187