Rate limits

Gridline limits how fast a company can call the API, and the limit follows the plan you pay for. It is a budget per minute: spend it however you like, and when it runs out you wait.

The budget

The budget belongs to the whole company, and it is spent by API keys: every key of the company draws from the same pool, so two busy scripts share it. A person using the dashboard is counted separately, at 300 requests a minute for each person, so using Gridline never uses up your API budget. A request from a visitor who is not signed in is counted per address, at 120 a minute.

Read your budget

Every answer carries three headers:

X-RateLimit-Limitoptional

The requests per minute you are being counted against: your plan's budget for an API key.

X-RateLimit-Remainingoptional

How many you have left in this window.

X-RateLimit-Resetoptional

When the window resets.

When you run out

Gridline answers 429 Too Many Requests with a Retry-After header and a message that tells you what to do:

{
  "statusCode": 429,
  "message": "Your company's free plan allows 30 requests per minute. Try again in 12 seconds. Upgrade to the basic plan (PATCH /subscriptions/me) for 120 requests per minute.",
  "correlationId": "…",
  "timestamp": "…"
}

• Wait the number of seconds in Retry-After, then continue.
• Spread requests out rather than sending bursts.
• Prefer webhooks or live updates to polling: they cost nothing against the budget.

Stricter limits on sensitive actions

A few endpoints have a small bucket of their own, per address (or per company when signed in), so they cannot be hammered:

These buckets do not draw from your main budget. That is on purpose: a company that has used up its requests can still upgrade its plan.