Watch a runaway loop
not be your problem.
Three simulated AI tasks share one API key. The third has a bug — it never stops. Toggle UsageWall on and off. See your bill rise calmly with the wall, or spike out of control without it. $0.05 per call, ~GPT-4o pricing.
Three tools, one API key, $40 monthly cap
All three call OpenAI through your single API key. The cap is $40 across the whole month — your dollar limit, not per-tool.
Every call is metered before it goes through
No SDK. No code change. You swap your provider key for a UsageWall key in any tool that takes one — Cursor, ChatGPT clients, agents, scripts. Same shape, we just count the tokens before they pass.
What you see in real time
No code. Two minutes. One number to remember: your monthly cap.
-
01
Paste your provider key
Your OpenAI, Anthropic, Google, or DeepSeek key. We encrypt it at rest. You can rotate or revoke any time — the key never leaves our infrastructure once you save it.
-
02
Set your monthly cap in dollars
One number. "Never more than $40 this month." Optional alerts at 50%, 80%, 100%. Resets the 1st of every month or on the date you choose.
-
03
Use the UsageWall key in your tools
Cursor, ChatGPT clients, agents, your own scripts — anywhere that takes
OPENAI_API_KEY. Same shape. Same models. We just count cents before each call goes through. -
04
You watch the meter. We pull the plug.
Real-time spend dashboard. The instant a call would push you over your cap, we return "budget reached" instead of letting your card pay. Hard cap, not a polite warning.