Cursor Rate Limit: Slow Pool or Service Outage? | Devdown
Cursor Rate Limit
A rate limit usually reflects request frequency, plan allowance, or a slow-pool queue—not a Cursor outage. Check live IDE, CLI, and Cloud Agents status before reviewing usage.
Error: rate limit exceeded. Please try again later.Operational
Check live status first, then make one targeted change and wait for the result. Repeated refreshes, sign-ins, or concurrent retries can worsen rate limits and queues. Follow the official incident when a component remains affected; when status is normal and only you are affected, record the error time, network, and account type before contacting support.
Use your browser to test public entry points. No account details or content are sent.
This tests public reachability from your current network, not login, quota, or private APIs.
These tools currently show as operational. Confirm they fit your workflow before switching.
Paste this Markdown into a README or internal wiki.
[](https://devdown.dev/status/cursor)Run this command to view current status.
curl -s https://devdown.dev/cliOperational
13:29:43 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗Operational
18:15:13 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗Operational
20:01:40 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗Operational
18:10:34 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
Operational
18:15:22 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗Operational
13:29:50 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗Operational
09:37:22 UTC
What are developers seeing?
Anonymous developer reports, aggregated over roughly 15 minutes. They are not official status.
No public incident
Official ↗24 HOUR TAPE
Each mark is a saved official snapshot. Blank means history is still accumulating, not operational.
Get an alert after an official status change.
Recent changes
Usually not. When official components are operational, allowance, concurrency, or a slow-pool queue is more likely.
It is queued capacity after plan limits. It does not always appear as an official service incident.
Troubleshooting
HTTP 529
The model is up, the queue is full. Official status can stay green. Wait or switch models. Your account is fine.
Claude Code 529
Rate limit
Claude 5-hour and weekly caps, Cursor slow pools — all look operational on status pages. Check Usage in the product, do not wait for this page to turn red.
Cursor Rate Limit
HTTP 429
HTTP 429 means a request was rate limited, not necessarily an outage. Check official status, retry later, and confirm quota and request rate.
/errors/429
Login
When Claude Code, Cursor, Codex, or GitHub Copilot login fails, check official Login status, your session, network, and regional restrictions first.
Login troubleshooting
DATA SOURCE
Status comes from official vendor pages and public-page availability observations, refreshed automatically.
Status records and official incidents are updated to help you understand service stability.
Source health · This refresh Request self-heal · 18:15:21 UTC