Rate limits
The actual numbers.
Two limits protect the API: your plan's monthly returned-job allowance and short request windows that keep traffic fair. There are no automatic overages.
Subscription allowances
Starter
100,000
jobs returned / month
Growth
300,000
jobs returned / month
Scale
1,000,000
jobs returned / month
Request rate limits
| surface | endpoints | limit |
|---|---|---|
| Job search | GET /v1/jobs/search | 60/minute |
| Job detail | GET /v1/jobs/{job_key} | 120/minute |
| Company directory | GET /v1/companies | 60/minute |
| Company detail | GET /v1/companies/{company_key} | 120/minute |
| Company jobs | GET /v1/companies/{company_key}/jobs | 60/minute |
| Feeds | GET /v1/feed/active · /count · /expired · /modified | 60/minute |
| Stats | GET /v1/stats | 120/minute |
| Sitemaps | GET /v1/sitemaps · /v1/sitemaps/{segment_id} | 120/minute |
| Health | GET /healthz | 120/minute |
Behavior at the limit
- Over the limit you get a
429problem+json response with aRetry-Afterheader — back off that many seconds and continue. - At the monthly allowance, quota-consuming requests are blocked with an actionable upgrade link. Praxy does not silently add overage charges.
- Successful responses carry
X-RateLimit-Limit,X-RateLimit-Remaining, andX-RateLimit-Resetheaders. - The MCP server is keyless and separate from these REST buckets.