AICostManager
Cost governance that can stop a runaway job, not just draw it a chart.
Most cost tools are dashboards: they tell you what you already spent. This one is built around enforcement. A single SDK call reports usage, the ingestion pipeline persists the raw event and returns immediately with a `triggered_limits` token, and hard limits come back as an encrypted JWT the client can verify locally — so a budget breach stops the next call at the caller, without a round trip to us and without trusting the network.
The provider catalogue synchronises from models.dev and normalises into reusable pricing templates, which is the unglamorous half: a cost tool whose price table is stale is a cost tool that lies. Usage rolls up asynchronously through Celery, exports run on a schedule with webhook delivery, and the whole thing is multi-tenant with per-team budgets and thresholds.
- Built with
- Django · Postgres · Celery · MCP
- Address
- aicostmanager.com
- Also answers for
- mcp.aicostmanager.com
- Status
- Live
More developer tools
AI Token Pricer
The pricing table we needed internally, published because everyone needs it.
Django · shares the AICostManager deployment
AI Prompt Bucket
Prompts treated as deployable artefacts, with environments and an audit trail.
Django · Postgres · Celery · MCP
ConfigPig
One config registry that hands each consumer the format it wants.
Django · Postgres · Celery · MCP