Skip to content

Eval usage

Your organization's own monthly eval count. This is what e10s meters for your account. It is not usage for your subscribers; e10s does not track that.

e10s is in early access and free to use. See e10s.io/pricing.

The early-access allotment

Every self-serve organization gets 250,000 evals a month. Eval responses warn at 80% and 100%. Seven days after you first reach 100% in a month, evals stop until the next UTC month. The rest of the API and the console keep working. If you need more, contact us.

If you have an arrangement with us, your limit may differ and evals may never stop. When a limit applies, the console home page shows it.

What counts

One eval is one 2xx response from either route:

  • POST …/subscribers/{subscriber_id}/entitlements/check
  • GET …/subscribers/{subscriber_id}/entitlements (?all=true is still one)

429, 409, and 5xx responses do not count. Catalog, subscriber, subscription, and override calls never count. Counts reset on the first of each calendar month, UTC.

Response headers

When a monthly allotment applies to your organization, both eval routes return:

Header Meaning
X-E10s-Eval-Limit Evals included this month
X-E10s-Eval-Used Evals counted so far this month (can lag a second or so)
X-E10s-Eval-Warn Only when warning: 80 (at 80% or more), 100 (at or over the limit), grace (over the limit, in or past the 7-day grace period, and evals will stop)

No headers when your organization has no monthly cap.

When evals stop

Seven days after your organization first reaches 100% in a month, both eval routes return 409 until the next UTC month:

{
  "error": {
    "type": "invalid_request_error",
    "code": "CONFLICT",
    "message": "Monthly eval allotment exceeded: 260,000 of 250,000 evals used for 2026-10. Evals resume next month; contact https://e10s.io/contact for more.",
    "request_id": "01H…",
    "details": { "field": "evals", "value": "2026-10" }
  }
}

This happens only when blocking is true in current usage. The rest of the API keeps working. The 100 requests per second burst limit (429) is separate.

Current usage

GET /org/usage/current

Scope: subscribers:read. The console home page shows the same numbers.

200

{
  "gated": true,
  "reason": null,
  "period": "2026-10",
  "count": 210000,
  "limit": 250000,
  "plan": "free",
  "blocking": true,
  "warn": "80",
  "allowed": true,
  "first_over_at": null,
  "blocked_at": null
}
Field Meaning
gated false = no monthly cap applies to your organization right now
reason When gated is false, why (for example unlimited). null when gated
period UTC month, YYYY-MM
count Evals counted this month
limit Evals included, or null
plan The key of the allotment that applies to you. Self-serve organizations report free; an arrangement with us may show a different key. null when none applies
blocking true = evals stop 7 days after first reaching 100%. false = evals keep answering past the limit
warn 80 | 100 | grace | null
allowed false once evals are stopped
first_over_at When this month's count first reached the limit, or null
blocked_at When evals stop (first_over_at + 7 days) if blocking is true, otherwise null