Skip to main content
This guide is not applicable if your organization is using a custom model provider.
Omni’s AI features consume credits each time they process a request. The Credits tab in AI Hub > Setup allows you to manage and monitor AI credit usage across your organization. In this guide, you’ll learn what credits are, what drives consumption, how to set limits that control AI usage and cost, and how to monitor usage across your organization.

​
Requirements

Organization Admin permissions are required to access Omni’s AI credit controls and monitoring tools.

​
What are credits?

Credits are the unit of measurement for AI usage. Every interaction with an AI feature in Omni — such as asking a question, generating a summary, asking AI to make data model changes, or using the MCP Server — consumes tokens from an LLM provider. Omni rolls those up into credits. Credit usage varies depending on several factors:
  • Task complexity — More complex questions, multi-step analysis, or larger datasets use more credits than simple lookups. Each message in a session carries prior context, so long-running sessions can compound.
  • Data model and context size — Omni sends context from the data model to the LLM to improve answer accuracy (field descriptions, ai_context, synonyms, etc.). Larger data models use more credits.
  • LLM model — More capable models (e.g., Sonnet-class) consume more credits per request than lighter models (e.g., Haiku-class). See AI model settings to configure tiers.

​
Managing credit usage

The Credits tab in AI Hub > Setup includes settings that allow you to limit AI credit consumption once usage reaches a threshold you define. If your organization has multiple Omni instances, these settings will apply to all instances. See Managing usage for multiple instances for more information.

​
Available controls

These settings can be used individually or together:
​
Set a hard cap on AI usage
Disables AI features when the specified credit threshold is reached. Overages are enabled by default; use this setting to limit them.
Requests that are processing when the cap is met will complete, but subsequent requests will be blocked. This means that the hard cap limit may be slightly exceeded to account for in-process requests.
​
Downgrade to a less expensive model
Switches AI features to a less expensive model (Claude Haiku) when the specified credit threshold is reached, reducing credit consumption.

​
Defining credit usage thresholds

To create a tiered approach, enable both the hard cap and downgrade settings. Set Downgrade to a less expensive model at a lower threshold than Set a hard cap on AI usage. As consumption rises, AI features switch to a less expensive model first, then stop entirely once the hard cap is reached.
To set a hard cap or less expensive model:
  1. Navigate to AI Hub > Setup.
  2. Click the Credits tab.
  3. Toggle the setting to on.
  4. In the credits field, enter the number of credits where usage should be hard-capped or downgraded.
  5. Click Save.
Once enabled, notifications will be automatically sent to Organization Admins when usage approaches and exceeds the thresholds.

​
Managing usage for multiple instances

Controls are set at the organization level. This means that if your organization has multiple Omni instances, these settings will apply to all instances in the organization. For example, organization A has two Omni instances: instance 1 and instance 2. If a control setting is set to 100 credits, the threshold will be triggered when usage from both instances is equal to 100 credits.

​
Monitoring credit usage

Omni provides two ways to monitor AI credit usage:
  • The Credits tab gives you an at-a-glance view of overall usage against your thresholds.
  • The Credit Tracking dashboard breaks usage down in detail, including over time, by feature, and by user.

​
Check usage at a glance

The Credits tab in AI Hub > Setup displays your organization’s total credit usage for the current month, plotted against your included credit allotment and any thresholds you’ve configured. This gives you a quick way to see how close you are to downgrading or reaching your hard cap.
Example usage visualization in the AI credits tab
The bar is marked with:
  • Included — the credit allotment included with your plan.
  • Downgrade and Cap — the thresholds you’ve set, if any.
Your current usage is labeled credits used this month and plotted along the bar, which is color-coded from green when you’re well within limits to red as usage approaches the cap.

​
Analyze usage in detail

The Credit Tracking dashboard, available in the Analytics section of Omni, helps you understand how credits are being used across your organization in greater detail. To access the dashboard, click Analytics in the left navigation of the Omni app and select the Credit Tracking dashboard. The dashboard displays credit consumption over time and can break down usage in various ways, such as by feature and user.

​
Optimizing credit usage

  • Adjust the LLM model tier. See AI model settings.
  • Trim unused fields, descriptions, or ai_context from the data model. Reducing context may lower per-call credit counts, but can increase the number of turns needed to get a good answer. The goal is optimizing efficiency overall.
  • If using Claude, use the ai_settings.conversation_prune_length parameter to control how aggressively conversation context is managed. When conversation context reaches the specified token threshold, older messages are automatically pruned from the conversation history to stay within limits.
See Tuning Omni AI for cost and quality for additional credit usage optimization tips.

​
Common questions

AI usage is calculated on a monthly basis and resets on the first of the month at 12AM UTC.
Omni’s AI is an agentic system. For a given prompt or request, it may fire off multiple tool calls to get the answer or complete the task.For example, asking “show me revenue by region” might involve the AI searching the data model for the right fields, building the query, creating a visualization, and summarizing results — each as a separate step. Each of these steps is logged individually in the tracking data, which is why a single user prompt can appear as several rows.
These are “tool calls” — intermediate steps where the AI determines which action to take next. They don’t have a user-facing prompt because they are part of the AI’s internal reasoning as it works through a request. See the question above for more context on why these occur.
Credit usage can vary between identical prompts due to differences in conversation context (earlier messages in the session) or the AI choosing a different path to arrive at the answer.
Omni caches AI summaries/sub-titles/descriptions so they will only recompute and use credits any time the data changes underneath (e.g., when a filter changes or it’s refreshed without cache).

​
Next steps