We use cache read (covering 90% of your work), which means you spend 10 times fewer tokens overall!
immediately after payment, you receive:
— api key
— base url
— link to your personal dashboard
other packages: 4m / 8m / 15m / 25m / 50m / 100m
works with all major tools
claude code, vs code, n8n, open cowork, pencil dev, opencode, python sdk, and any application or backend that supports standard api connections.
important: prompt-cache savings
cached tokens are charged 90% cheaper
as a result, the package or balance provides approximately 5–6 times more total usage.
frequently asked questions
are there any speed limits?
up to 120 requests per minute per api key. actual response speed depends on the selected model, context size, and current load.