KeyGoodZero-Compromise LLM Budget Firewall
Control your AI API spending.
Create virtual API keys with hard spending limits. When the budget is gone, requests stop — before they hit OpenAI.
OpenAI-compatible · Bring your own key · Streaming supported
AI API costs can get out of control.
Shared master keys
You shouldn’t hand out your OpenAI key to every agent or teammate.
No hard stop
A loop or runaway agent can keep calling the API until the bill hurts.
Weak control
Limits and allowlists should be enforced — not just reported after the fact.
How it works
No custom SDK required. Use your existing OpenAI client with a different base URL.
- 01
Connect
Bind your OpenAI API key. It is encrypted at rest.
- 02
Create
Issue a virtual key with daily/monthly limits, model allowlist, and concurrency.
- 03
Use
Send OpenAI-compatible requests to KeyGood with the virtual key.
App → Virtual Key → KeyGood → OpenAI
↑
hard limitsWhat you get in V1
Virtual keys
Isolated keys; full secret shown once; only a hash is stored.
Hard spending limits
Requests stop when the budget is exhausted (HTTP 402).
Budget reservation
Budget is reserved before the upstream call.
Usage tracking
See spend and recent requests in the dashboard.
Model allowlist
Restrict which models a key may call.
Email alerts
Get notified near and at budget limits.
Simple pricing
Early access while we validate Hard Stop with real users. No complex tiers on day one.
View PricingBuilt around your keys
- ·Master API keys encrypted at rest (AES-256-GCM)
- ·Virtual keys stored as irreversible hashes — not recoverable after creation
- ·Secrets are handled on the server; not exposed in the browser dashboard
Take control of your AI API spending.
Bring your own OpenAI key. Create virtual keys. Set hard limits.
Get Started