Grafana Assistant pricing
Grafana Assistant bills on two dimensions: active AI users and token usage. Charges are calculated for each Grafana Cloud organization and billing month.
For current rates and included usage, refer to the pricing page.
How Assistant is billed
Assistant charges appear on your invoice as two line items:
- Assistant covers active AI users above the number included with your plan.
- AI Tokens covers user and system-initiated token usage above the amounts included with your plan.
The Grafana Cloud Pro platform fee is a separate charge and includes the first three active AI users in the organization.
Assistant billing is independent of other Grafana Cloud features such as Visualization or IRM. If you use Assistant from a self-managed Grafana deployment, usage is still tracked against the paired Grafana Cloud stack. The LLM plugin is open source and has no charge.
Definitions
Active AI user: A user who does any of the following during the billing month:
- Sends a message to Assistant
- Uses an Assistant action, for example Explain in Assistant
- Uses Assistant in Grafana, Workspace, Slack, Microsoft Teams, or the
gcxCLI - Connects through the Grafana Cloud MCP server
- Runs an Assistant Automation manually, or has one run automatically
An active AI user is counted once per month across all stacks in the organization.
Token: The basic unit of content processed by an AI model. Token usage is counted across every Assistant surface, including integrations, Slack, Microsoft Teams, API integrations, and the gcx CLI.
User token usage: Usage attributed to an individual, covering Assistant use in Grafana, Workspace, Slack, Microsoft Teams, and the gcx CLI, features invoked through the Grafana Cloud MCP server, and Automation runs. A manual run counts toward the user who starts it. Scheduled and event-triggered runs count toward the creator.
System-initiated token usage: Assistant API requests authenticated with a Grafana service account, and investigations triggered by automated systems. These don’t count as an active AI user and don’t receive the per-user token allowance. If a user manually interacts with a system-initiated session, that interaction counts toward the user’s own token usage.
How usage is calculated
Each active AI user on a Free or paid plan has 40 million included tokens for the billing month. Included tokens aren’t shared between users and unused tokens don’t carry over. Only a user’s usage above their own allowance is charged, and one user’s unused tokens don’t offset another user’s usage.
Free and Pro plans include 25 million tokens for system-initiated usage per organization each billing month. This allowance is shared across all stacks in the organization, and only usage above it is charged.
Connecting through the Grafana Cloud MCP server counts as active use even if you don’t use any Grafana AI tokens. Querying Grafana data through the MCP server doesn’t consume tokens, but invoking an Assistant feature that uses a Grafana-provided AI model does.
Special considerations
Assistant Investigations
Assistant and Assistant Investigations are separate products with their own pricing, but both are billed in tokens and draw from the same per-user allowance and system-initiated token pool.
Pricing and metering for Assistant Investigations start on October 1, 2026. Before that date, investigation token usage appears in your usage views but doesn’t count toward allowances and isn’t billed.
Manually triggered investigations count as user-initiated usage. Investigations triggered by automated systems count toward the system-initiated pool until a user interacts with them.
Watchers
Assistant Watchers are free during the preview period. Preview usage may be subject to usage caps, and pricing and metering may change when Watchers become generally available.
Grafana Labs support usage
When Grafana Labs employees work on your stack on behalf of your organization, their Assistant activity doesn’t count toward active AI users or your stack’s token usage.
Contracted accounts
Contracted accounts may have different prices or included usage, and Assistant and Assistant Investigations are billed separately. Refer to your contract or contact your Grafana Labs account team for the terms that apply to your organization.
Controlling usage
Administrators can cap Assistant consumption by setting monthly active user and token limits, or by restricting access with RBAC. On Free plans, limits are enforced as hard limits. For how to configure them, refer to Understand Grafana Assistant pricing and usage.
View your usage
- In your Grafana Cloud stack, go to Cost Management and Billing > Usage.
- Select Assistant to review token usage by user and service account, and to compare usage against what’s included with your plan.
For usage and limits scoped to a single stack, go to Assistant > Usage. Refer to Understand Grafana Assistant usage and cost.
