Vercel Connect Limits
The following limits apply. Contact your account team if you need higher limits.
| Resource | Limit |
|---|---|
| Trigger destinations per connector | 3 |
Projects returned by ?include=projects | 100 (paginated beyond) |
| Default token validity buffer | 30 seconds |
| Connector branding icon | PNG or JPEG, square |
Vercel Connect applies per-minute rate limits to protect shared infrastructure. Each limit resets after one minute.
Limits are scoped by team, IP address, or connector ID, depending on how the endpoint is accessed.
SDK calls and CLI commands authenticate with your team identity (Vercel OIDC token, access token, or session). Rate limits are scoped per team.
| Operation | Interface | Description |
|---|---|---|
getToken | SDK | Request a provider token |
getTokenResponse | SDK | Request a provider token with full response metadata |
getConnectorMetadata | SDK | Retrieve a connector's public metadata |
vercel connect list | CLI | List connectors for your team or project |
vercel connect token | CLI | Get a runtime token from a connector |
Dashboard reads (viewing connectors, project links, installations, observability events, and metrics) also count against this limit.
| Operation | Interface | Description |
|---|---|---|
revokeToken | SDK | Revoke a provider grant and clear the token cache |
vercel connect create | CLI | Create a new connector |
vercel connect attach | CLI | Link a project to a connector |
vercel connect detach | CLI | Unlink a project from a connector |
vercel connect update | CLI | Update connector branding |
vercel connect remove | CLI | Delete a connector |
Manual Vercel Connect Dashboard write operations (creating, updating, or deleting connectors, managing project links, and starting authorization or installation flows) also count against these limits.
These endpoints are called by external OAuth clients against Vercel Connect acting as an OAuth authorization server. They do not require Vercel user authentication. Each endpoint is rate-limited to 6,000 requests per minute, keyed as shown.
| Endpoint | Method | Keyed by | Description |
|---|---|---|---|
/connect/oauth/authorize | GET | IP & Connector | Start OAuth authorization |
/connect/oauth/token | POST | Connector | Exchange authorization code for tokens |
/connect/oauth/revoke | POST | Connector | Revoke a token issued by Connect to an OAuth client |
/connect/oauth/userinfo | GET | Connector | Retrieve user claims for an access token |
| Endpoint | Method | Keyed by | Description |
|---|---|---|---|
/connect/trigger/:connectorId | POST | Connector | Receive a webhook from a provider |
When you exceed a rate limit, the API returns a 429 Too Many Requests response. Wait for the limit window to reset (one minute) before retrying.
For the SDK methods getToken and getTokenResponse, a 429 response throws an error in your application. Use in-process token caching to reduce the number of token requests and stay within limits.
Was this helpful?