Anthropic 429 enforced_spend_limit_reached: the spend-cap you should not retry

The short version

• Anthropic's usage tiers carry a monthly spend cap. Hitting it pauses API usage until 00:00 UTC on the 1st of next month (or until you get a higher tier).
• While paused, requests return 429 with error.type = rate_limit_error and error.details.error_code = enforced_spend_limit_reached.
• There is no retry-after header. Do not tight-loop or rely on the SDK's default retries.
• A spend limit you set is a different path: HTTP 400 invalid_request_error, message starting with “You have reached your specified…”.
• Ordinary rate limits are also 429, but they usually carry retry-after and do not set that error_code.

1. What the response looks like

Anthropic's rate limits docs (read 7 Nov 2026) show this shape when the tier monthly spend cap is reached:

{
  "type": "error",
  "error": {
    "type": "rate_limit_error",
    "message": "You have reached your API usage limits: your organization has crossed its monthly API usage threshold, set based on your organization's API tier. You will regain access on 2026-09-01 at 00:00 UTC.",
    "details": { "error_code": "enforced_spend_limit_reached" }
  },
  "request_id": "req_018EeWyXxfu5pfWkrYcMdjWG"
}

On the Messages API, read error.details.error_code. The documented sample message also names the UTC date when access resumes. The errors page groups this under 429 rate_limit_error along with ordinary rate limits and Claude Code workspace spend limits — so status alone is not enough.

2. Triage table: which “limit” did you hit?

3. Current tier caps (and where to see yours)

Each of the Start, Build and Scale tiers carries a monthly spend cap — the maximum your organization can spend on the API each calendar month (docs, 7 Nov 2026):

You can see your organization's tier and current limits on the Rate limits page in the Claude Console, and you can set your own lower spend limit on the Billing page. Your own limit cannot exceed the tier cap. The Default Workspace still cannot have a spend limit — see Anthropic Default Workspace limits.

4. Handle it in code (don't treat every 429 alike)

async function createMessage(body) {
  const res = await fetch("https://api.anthropic.com/v1/messages", {
    method: "POST",
    headers: {
      "x-api-key": process.env.ANTHROPIC_API_KEY,
      "anthropic-version": "2023-06-01",
      "content-type": "application/json",
    },
    body: JSON.stringify(body),
  });
  if (res.status !== 429) return res;

  const payload = await res.json().catch(() => ({}));
  const code = payload?.error?.details?.error_code;
  if (code === "enforced_spend_limit_reached") {
    // No retry-after. Alert a human; do not tight-loop.
    throw new Error("Anthropic tier monthly spend cap reached");
  }
  const wait = Number(res.headers.get("retry-after"));
  if (Number.isFinite(wait) && wait > 0) {
    await new Promise((r) => setTimeout(r, wait * 1000));
    return fetch("https://api.anthropic.com/v1/messages", {
      method: "POST",
      headers: {
        "x-api-key": process.env.ANTHROPIC_API_KEY,
        "anthropic-version": "2023-06-01",
        "content-type": "application/json",
      },
      body: JSON.stringify(body),
    });
  }
  throw new Error("Anthropic 429 without a clear retry");
}

The official SDK retries transient failures (including rate limits) with exponential backoff by default and honours retry-after when present. Because the spend-cap 429 has no that header, those retries keep failing until the month rolls or the tier rises — so catch enforced_spend_limit_reached explicitly and stop.

5. What to do when production hits the cap

1. Confirm error.details.error_code === "enforced_spend_limit_reached" (and note the resume date in the message).
2. Stop automatic retries for that organization; page a human.
3. Open the Rate limits page and use Request tier increase if you need access before next month (Claude Platform on AWS: contact your account rep or support; that flow is not available).
4. Separately, set your own spend limits below the tier cap so the next surprise is a 400 you chose, not a hard stop at the tier ceiling. Remember: no spend limit on the Default Workspace — move agent keys to a capped workspace.
5. Something outside the app that tells you the same day when spend is far above normal, before the monthly cap trips.

Not sure whether your Anthropic (and OpenAI / OpenRouter) limit settings would actually stop a loop or a leaked key? The free AI key limit audit checks them in a few minutes (no sign-up, no key needed). Related: credit balance too low, workspace spend limits via the API, runaway AI agent cost.

Where KeyBark fits (and where it doesn't)

KeyBark reads Anthropic organization cost from our servers about every 15 minutes during the beta (Admin cost report, including Default Workspace spend) and emails you when a daily or monthly dollar threshold you set is crossed, or when today's spend jumps above your 7-day average. That covers the “spend is far above normal” item before the monthly tier cap trips. On paid plans an opt-in kill switch (beta) can disable a named key when a threshold is crossed.

Limits, honestly: KeyBark does not see 429 responses from your app, does not read your usage tier or monthly spend cap, and does not request tier increases for you. Checks run on a schedule, so an alert can arrive after money has been spent. KeyBark is not a proxy and can't block a request. Set Anthropic's own spend limits (and a sensible tier) first; KeyBark is a second line of warning on top.

FAQ

Sources (read 7 Nov 2026): Anthropic Rate limits and API errors. KeyBark is not affiliated with Anthropic.