API and AI agent usage
Everre has no seats and no subscription. You pay for what you use, and that includes the API — whether you are calling it from your own script, a tool like Zapier, or an AI agent working on your behalf.
What counts
A billable request is any call to the Everre API made with an API key or a personal access token, that we answered successfully.
Using the Everre web app or mobile app costs nothing. Those are signed-in sessions, not API keys, and they are never metered — clicking around the product is free no matter how much of it you do.
These are free:
- Anything you do in the Everre app itself.
- Requests we rejected: a bad parameter, a missing permission, a record that does not exist.
- Requests we could not answer because of an error on our side.
- Requests refused for going too fast (rate limiting).
You are only charged for work we actually did for you.
What it costs
| Included every month | 5,000 requests, no card required |
| After that | $3.00 per 1,000 requests — $0.003 each |
The allowance resets at the start of each calendar month and is shared across your whole organisation.
To put that in perspective:
- A nightly sync that pulls a few hundred records: free, permanently.
- An AI assistant answering questions about your pipeline through the day: usually free.
- An agent doing sustained research and data entry all day, every working day: around $120 a month.
AI agents and connectors
If you connect an AI agent to Everre — your own, or a connector we publish — it authenticates with a personal access token, so its calls are metered exactly like any other API traffic. There is no separate charge for using an agent, and no markup for connecting one you built yourself.
Note that this is separate from Everre's own AI features. If your agent asks Everre to generate or analyse something, that AI work is billed on the AI meter as well, because it costs us model time. Reading and writing your own records only touches the API meter.
Seeing your usage
Settings → Access & Usage shows the month to date, broken down by meter. API requests appear as their own tile with the request count and what it has cost, alongside AI, signatures, storage, websites and email.
Counts are written in batches rather than one at a time, so the figure can lag real time by up to a minute. Nothing is lost — it catches up.
Spend caps
The caps you set in Access & Usage cover total spend across every meter, and they apply to API usage too.
When an organisation reaches its cap, its API keys and personal access tokens
stop working. Requests are refused with a 429 and the code
AI_LIMIT_REACHED. Anything running against the API — scripts, integrations,
agents — stops until the cap is raised.
The Everre app keeps working. Signing in, opening Access & Usage and raising the cap are all unaffected, so reaching a cap never locks you out of the screen where you fix it. An administrator can raise or remove the cap and API access resumes immediately.
If you have passed the free allowance and have no card on file, requests are
refused with a 402 and the code PAYMENT_METHOD_REQUIRED. Add a card in
Settings → Billing to continue.
Two different 429s
Both rate limiting and spend caps answer with HTTP 429, so check error.code
to know which happened — the fix is different:
| Code | Meaning | What to do |
|---|---|---|
RATE_LIMITED | Too many requests too quickly | Back off and retry; see Retry-After |
AI_LIMIT_REACHED | The monthly spend cap is reached | An administrator raises the cap |
Rate limits and spend caps are separate controls. Staying under your rate limit does not mean you are under your spend cap, and vice versa.