Text of docs.gitlab.com › administration › instance_limits as Styx crawled it on Nov 19, 2026. The page may have changed since. Only its words are kept, without pictures or layout, and only the first part of a long page.

Open the live page

GitLab application limits | GitLab Docs

/ GitLab application limits Tier : Free, Premium, Ultimate Offering : GitLab Self-Managed GitLab, like most large applications, enforces limits in certain features to maintain a minimum quality of performance. Allowing some features to be limitless could affect security, performance, data, or could even exhaust the allocated resources for the application. Instance configuration In the instance configuration page, you can find information about some of the settings that are used in your current GitLab instance.

Depending on which limits you have configured, you can see: SSH host keys information CI/CD limits GitLab Pages limits Package registry limits Rate limits Size limits Because this page is visible to everybody, unauthenticated users only see the information that is relevant to them. To visit the instance configuration page: In the left sidebar, select Help ( ) > Help . On the Help page, select Check the current instance configuration.

The direct URL is <gitlab_url>/help/instance_configuration . For GitLab.com, you can visit https://gitlab.com/help/instance_configuration . Rate limits Rate limits can be used to improve the security and durability of GitLab. For a complete list of rate limits, and how to change each one, see rate limits . By protected path This setting limits the request rate on specific paths. GitLab rate limits the following paths for POST requests by default: '/users/password', '/users/sign_in', '/api/#{API::API.version}/session.json', '/api/#{API::API.version}/session', '/users', '/users/confirmation', '/unsubscribes/', '/import/github/personal_access_token', '/admin/session' GitLab rate limits the following paths for GET requests by default: '/users/sign_in_path' Read more about protected path rate limits .

Default rate limit : After 10 requests, the client must wait 60 seconds before trying again. Import and export These settings limit file imports and exports for groups and projects, and imports from other sources. Limit Default (per minute per user) Project import 6 import requests Project export 6 export requests Project export download 1 download request Group import 6 import requests Group export 6 export requests Group export download 1 download request GitHub gists import 6 import requests These settings can be configured .

Direct transfer migration The following limits apply on migration by direct transfer. Limit Default Configurable Number of migrations by a destination GitLab instance per minute per user. 6 No Relations export requests per minute per user for each project or group. 6 Relations export downloads per minute per user for each project or group. 100 Time to wait for decompressing an archive file. 210 seconds No Length of an NDJSON row.

50 MB No Time until an empty export status on source instance is raised. 5 minutes No Relation size that can be downloaded from the source instance. 5 GiB Size of a decompressed archive. 10 GiB For more information on changing configurable limits, see import and export settings and import and export rate limits . Member Invitations Limit the maximum daily member invitations allowed per group hierarchy. GitLab.com: Free members may invite 20 members per day, Premium trial and Ultimate trial members may invite 50 members per day.

GitLab Self-Managed: Invites are not limited. Webhook rate limit Limit the number of times each minute that webhooks in a top-level namespace can be called. All project and group webhooks in the namespace share this limit. System hooks managed by organization administrators use the same limit, counted for each organization. All system hooks in an organization share this limit. System hooks managed by instance administrators are not rate limited.

Calls that exceed the rate limit are logged in auth.log . To set this limit for a GitLab Self-Managed instance, use the Plan Limits API or run the following in the GitLab Rails console : ruby Plan . default . actual_limits . update! ( web_hook_calls : 10 ) Set the limit to 0 to disable it. Default rate limit : Disabled (unlimited). Integration execution rate limit Introduced as a beta in GitLab 19.6 with a feature flag named integration_execution_budget .

Disabled by default. This setting is in beta and might be changed or removed without notice. Limit the number of times each minute that integration events in a top-level namespace are enqueued for execution. All project and group integrations in the namespace share this limit, including integrations inherited from an instance-level or group-level integration. Datadog has its own, separate limit, because it receives every pipeline and job event.

The limits are: integration_executions : all integrations except Datadog. integration_executions_datadog : Datadog only. Events that exceed a limit are dropped and are not retried. To set these limits for a GitLab Self-Managed instance, use the Plan Limits API or run the following in the GitLab Rails console : ruby Plan . default . actual_limits . update! ( integration_executions : 500 , integration_executions_datadog : 3000 ) Set a limit to 0 to disable it.

By default, the feature only counts events that exceed a limit and logs them, without dropping anything. Events are dropped only when the enforce_integration_execution_budget (or, for Datadog, enforce_integration_execution_budget_datadog ) feature flag is also enabled. Default rate limit : Disabled (unlimited). Service Desk email rate limit Introduced in GitLab 19.4 with a feature flag named service_desk_email_rate_limit .

Disabled by default. The availability of this feature is controlled by a feature flag. For more information, see the history. Limit the number of outbound Service Desk notification emails that a top-level namespace can send per hour and per day. This includes the ticket creation confirmation email, new comment notification emails, and new participant notification emails. All projects and subgroups in the namespace share these limits.

Each email sent counts as one increment, so a comment that notifies 10 participants counts as 10 emails. When a namespace reaches either limit, GitLab suppresses the outbound email but still creates the ticket, processes the comment, or adds the participant. GitLab adds an internal note to the affected ticket to record that email sending was suppressed, at most once per ticket per hour. To set these limits for a GitLab Self-Managed instance, use the Plan Limits API or run the following in the GitLab Rails console : ruby Plan . default . actual_limits . update! ( service_desk_outbound_emails_per_hour : 100 , service_desk_outbound_emails_per_day : 1000 ) Set a limit to 0 to disable it.

Default rate limit : Disabled (unlimited) on GitLab Self-Managed. For GitLab.com plan-specific limits, see Rate limits on GitLab.com . Search rate limit This setting limits search requests as follows: Limit Default (requests per minute) Authenticated user 30 Unauthenticated user 10 Search requests that exceed the search rate limit per minute return the following error: This endpoint has been requested too many times. Try again later.

Autocomplete users rate limit Introduced in GitLab 17.10 with a feature flag named autocomplete_users_rate_limit . Disabled by default. Generally available in GitLab 18.1. Feature flag autocomplete_users_rate_limit removed. This setting limits autocomplete users requests as follows: Limit Default (requests per minute) Authenticated user 300 Unauthenticated user 100 Autocomplete requests that exceed the autocomplete rate limit per minute return the following error: This endpoint has been requested too many times.

Try again later. Dependency proxy rate limit Authenticated requests to the dependency proxy for container images can use a dedicated rate limit instead of the authenticated web request rate limit. Default rate limit : Disabled. When enabled, defaults to 1,000 requests per 15 seconds per user. MCP server rate limit Authenticated requests to the GitLab MCP server ( POST /api/v4/mcp ) can use a dedi

The rest of this page is on the live site.