Documentation for automated readers
A curated documentation index is available at: https://grafana.com/llms.txt
A complete documentation index is available at: https://grafana.com/llms-full.txt
These indexes can help with page discovery before fetching individual documents.
This page is also available in Markdown, which may be easier for automated readers and AI tools to parse than HTML. The Markdown version is available at https://grafana.com/docs/grafana-cloud/platform/grafana-assistant/get-started/self-managed.md, or by sending Accept: text/markdown to https://grafana.com/docs/grafana-cloud/platform/grafana-assistant/get-started/self-managed/. For broader documentation discovery, the curated index is available at https://grafana.com/llms.txt and the complete index is available at https://grafana.com/llms-full.txt.
Set up Grafana Assistant in self-managed Grafana
Install the Assistant app in a self-managed Grafana deployment and connect it to a Grafana Cloud stack. This article explains how it works, what you need before you start, and which features are available.
Understand how it works
Assistant uses a hybrid deployment model for self-managed installations.
The Assistant UI runs in your self-managed Grafana deployment. The backend, usage limits, and billing stay in the Grafana Cloud stack you connect during setup. If you run Grafana OSS or Grafana Enterprise, this is the supported way to use Assistant outside Grafana Cloud.
This means the self-managed experience doesn’t include every Grafana Cloud feature. Assistant keeps the core chat and dashboarding workflows, but some backend-dependent features remain available only in Grafana Cloud.
Understand what data goes to Grafana Cloud
When you use Assistant, the prompts you submit and the query context needed to generate a response are sent to the Grafana Cloud backend for processing. Your underlying data sources and infrastructure remain in your own environment. Grafana Cloud processes requests in the region associated with your connected stack.
For full details on data handling, refer to Manage your data privacy and security.
Before you begin
- A Grafana Cloud stack where Assistant is available or enabled
- A self-managed Grafana 13.0.0+ deployment
- Organization administrator access in the self-managed Grafana deployment
- Administrator access in the Grafana Cloud stack that will back the deployment
Connect your deployment
- Make sure Grafana Assistant is available or enabled in your Grafana Cloud stack.
- Sign in to your self-managed Grafana instance as an organization administrator.
- Open Administration > Plugins and data > Plugins.
- Search for Grafana Assistant or go directly to
/plugins/grafana-assistant-app. - Click Install.
- Open Connection.
- Click Connect to Grafana Cloud and complete the authorization flow.
After the connection succeeds, the page reloads and Assistant becomes available in the self-managed Grafana UI.
To configure the connection manually, expand Manual configuration in Connection and enter the following values from your Grafana Cloud stack:
- Backend URL
- Instance ID
- API Token
Note
Connect to Grafana Cloud sends your Grafana URL, including internal hostnames, to Grafana Cloud so the two deployments can be paired.
Review available features
Assistant in self-managed deployments keeps the main chat workflows available and hides features that depend on the full Grafana Cloud backend.
Available in self-managed deployments:
- Chat and prompt-based assistance
- Dashboard creation and editing
- Query generation and explanation
- Navigation across Grafana resources
- Rules, quickstarts, and skills
- MCP server integrations
Not available in self-managed deployments:
- Investigations and investigation memory
- Slack integration
- Infrastructure memory
- Grafana Cloud MCP connections
- CLI auth tokens
- SQL table discovery
- Automations and sandbox settings
- Anonymous access to the Assistant app
Understand pricing and limits
Self-managed deployments use the same usage limits and pricing as Assistant in Grafana Cloud. Limits apply per connected Grafana Cloud stack.
Refer to Grafana Assistant pricing to see the current tiers, included seats, and overage rates.
Note
If you’re evaluating Assistant for a large organization, contact your Grafana account team for a tailored plan.
Manage access
Assistant uses Grafana RBAC in the Grafana deployment where the plugin runs.
For Grafana Enterprise deployments where exact RBAC is enabled, Assistant honors the permissions granted by basic, custom, and Assistant-specific roles. The plugin resolves those permissions through your local Grafana instance, so the Grafana Cloud backend doesn’t need network access to it. Contact your Grafana account team to confirm availability.
Other self-managed deployments, including deployments without RBAC, use a compatibility fallback based on the user’s basic organization role. Viewers can access and read Assistant resources, Editors can also create, update, and delete non-administrative resources, and Admins receive administrative permissions. Custom and Assistant-specific roles can’t grant permissions beyond this fallback.
Use plugins.app:access scoped to plugins:id:grafana-assistant-app to control who can open Assistant in your self-managed deployment. Permissions alone don’t enable features that aren’t available in self-managed deployments.
Configure internal authorization access
For exact RBAC, the Assistant backend plugin calls the Grafana API inside your deployment. By default, it uses http://127.0.0.1:3000 and preserves any subpath from your Grafana root_url. This avoids routing authorization requests through a public ingress or load balancer.
If Grafana uses a custom internal address, port, protocol, or certificate, configure an internal URL in grafana.ini:
[plugin.grafana-assistant-app]
grafana_internal_url = http://grafana.monitoring.svc.cluster.local:3000The URL must be reachable from the Grafana pod or host. It can include a subpath, but it must not include credentials, a query, or a fragment. Restart Grafana after changing this setting.
Usage and limits are tracked against the connected Grafana Cloud stack. To review or change your plan, refer to Grafana Assistant pricing.
Next steps
Was this page helpful?
Related resources from Grafana Labs


