Reference
Plans & limits
Each plan sets per-key rate limits (requests per minute and per day) and account-wide monthly quotas (words and requests). The numbers below come from the same plan configuration the API enforces. New accounts start on Starter. To move to Growth or Scale, send a request from the Plan & access page of your dashboard.
Starter
$0/mo
- 1 req/min per key
- 3 req/day per key
- 90 requests/mo
- 20,000 words/mo
- 1 concurrent job
Growth
$49/mo
- 15 req/min per key
- 500 req/day per key
- 5,000 requests/mo
- 200,000 words/mo
- 3 concurrent jobs
Scale
$199/mo
- 60 req/min per key
- 5,000 req/day per key
- 50,000 requests/mo
- 1,200,000 words/mo
- 10 concurrent jobs
Rate limits & quotas
| Plan | Requests/min (per key) | Requests/day (per key) | Requests/mo (account) | Words/mo (account) | Concurrent jobs |
|---|---|---|---|---|---|
| Starter | 1 | 3 | 90 | 20,000 | 1 |
| Growth | 15 | 500 | 5,000 | 200,000 | 3 |
| Scale | 60 | 5,000 | 50,000 | 1,200,000 | 10 |
How each limit works
- Requests per minute and per day apply to each API key on
POST /v1/generateandPOST /v1/jobs. The day window resets at UTC midnight and the minute window at the next full minute. Over the limit, you get429 RATE_LIMITEDwith aRetry-Afterheader. Every request that reaches this check counts, including ones that later fail validation or generation. On Starter, the daily limit is usually the first one you hit. - Monthly words and monthly requests apply to your whole account (all keys combined) over the calendar month in UTC. Only successful generations count: the request, plus the words in the returned article. Once either total reaches the plan allowance, further generation requests return
429 QUOTA_EXCEEDEDuntil the next month starts. The check runs before a request starts, so a single article is never cut short. The generation that crosses the limit completes in full. - Concurrent jobs is how many of your
POST /v1/jobsjobs are processed at the same time. Additional jobs wait in line rather than failing. - No per-request word cap. Article length isn’t limited or targeted, and
constraints.maxWordsdoesn’t change it.GET /v1/usagestill reports alimits.maxWordsPerRequestvalue, but it is not enforced. GET /v1/usageandGET /v1/jobs/{jobId}don’t count toward any limit. Check your headroom with GET /v1/usage.
Rate limited response
429
{
"error": {
"code": "RATE_LIMITED",
"message": "Too many requests per minute. Please slow down.",
"requestId": "req_11223344aa",
"details": {
"limit": 15,
"remaining": 0,
"resetAt": "2026-09-14T10:21:00.000Z"
}
}
}Quota exceeded response
Once consumed.words reaches limits.monthlyWords, or consumed.requests reaches the monthly request allowance, generation requests return the response below. The message names the quota that was reached (word or request), and details carries both totals.
429
{
"error": {
"code": "QUOTA_EXCEEDED",
"message": "Monthly word quota exceeded for plan 'growth'. Upgrade your plan or wait for the next billing period.",
"requestId": "req_f6e5d4c3b2",
"details": {
"monthlyWords": 200000,
"consumedWords": 200000,
"monthlyRequests": 5000,
"consumedRequests": 231,
"periodEnd": "2026-10-01T00:00:00.000Z"
}
}
}