What is API usage billing?
Updated · By Robert Breen
API usage billing means paying for an AI model by how much you use it, typically by the tokens your requests send and receive, rather than a flat monthly subscription. Developer platforms like the OpenAI API work this way, and it is what powers AI inside tools like n8n.
Why it matters for a small business
The first surprise for many business owners is that a ChatGPT subscription does not cover automations. An n8n workflow talks to the model through the API, which has its own account, its own balance and its own bill. Knowing that up front saves the confusion of an agent that throws errors even though "we already pay for ChatGPT".
Usage billing is good news for small automations: a few hundred short requests a month can cost very little, and you pay nothing in a month the workflow is idle. The flip side is that cost scales with use. A workflow that runs on every email, uses a large model or sends long prompts can add up, so it is worth setting a budget and checking usage once a workflow is live.
In a real lesson: Build an AI Follow-Up Agent for HVAC Service Calls
The HVAC follow-up agent lesson walks through the billing side before any building. To give an n8n agent for Cedar Ridge Heating & Air (a made-up HVAC company) a model, you go to platform.openai.com → Billing and click Add to credit balance. The voice-over stresses that this is the developer side, not ChatGPT, that it bills separately, and that a brand new key with a zero balance just returns errors.
The lesson suggests ten dollars and explains why that goes far: you are charged per request, not per month, and each batch of follow-ups costs fractions of a cent. Then you click API Keys, + Create new secret key, name it n8n agent, and copy it once.
Later, under Credential to connect with, you paste the key into the OpenAI Chat Model and click Save, and pick gpt-5-mini. Every run that writes follow-ups for Linda Morales's furnace tune-up or Grace Chen's heat pump install now draws a little from that balance.
Try this lesson free or read the step-by-step guide.
Common confusions
API billing vs a ChatGPT subscription
A ChatGPT plan pays for a person using the chat app. API usage is billed separately, per request, for software like n8n. Having one does not include the other. See OpenAI API for the setup side.
Usage billing vs n8n's own costs
The OpenAI balance pays for the model. Running n8n itself is a separate cost, whether on n8n Cloud or on a server you host. Budget for both.
Tips
- Add credit before creating the key, or the key will only return errors.
- Set a monthly budget or usage limit in the provider's billing settings if one is offered.
- Check the provider's current pricing page instead of relying on old numbers; rates change.
- Name each key after what uses it so you can trace spending and revoke one without breaking others.
Related terms
Where you use it: free lessons
- Build an AI Follow-Up Agent for HVAC Service Calls (n8n, 12 min)
- n8n AI Agent Tutorial: Save Social Media Ideas to Google Sheets (n8n, 12 min)
- AI Receipt Extractor: Receipts to Google Sheets with n8n (n8n, 12 min)
Prompt templates that use it
Frequently asked questions
- Does ChatGPT Plus include API usage?
- No. ChatGPT plans and the OpenAI API are billed separately. The API needs its own credit balance at platform.openai.com.
- How do I estimate API costs?
- Run the workflow on a handful of real examples, then check the usage page in your provider's dashboard. Multiply by your expected monthly volume and add a margin.
- What is a good model choice to keep costs down?
- Use a small, fast model for short text jobs, as the lessons do with gpt-5-mini, and switch to a more capable one only when a task needs it.