Usage and Cost#

Open Analytics → Usage and Cost to review AI consumption, spend, and reliability across the organization. Break down request volume, tokens, and cost by model, user, team, or agent, and watch whether trends stay within budget and health targets. For scenario-based examples, see Use Cases — Usage and Cost.

Use this page when you need to:

• Monitor AI adoption via requests and token consumption.
• Analyze spending across models, users, teams, and agents.
• Measure reliability (success rate, failures, provider latency).
• Spot unusual usage or spend before it becomes costly.
• Support budgeting and capacity planning with trend data.

From this page you can move from organization-wide totals to cost drivers and activity for a selected model, user, agent, or team.

Next: start on Organization totals, dig into Cost breakdown when spend shifts, then use activity tabs for the dimension you care about. Open Traces for a single request, Optimizations for savings, or Dashboards for pinned monitoring.

The page includes these views:

• Totals — Overall usage, volume, tokens, and health. See Organization totals.
• Cost — Spend by model and user. See Cost breakdown.
• Model Activity — Analytics for a selected model. See Model Activity.
• User Activity — Analytics for an individual user. See User Activity.
• Agent Activity — Analytics for an application or agent. See Agent Activity.
• Team Activity — Department-level adoption and budgeting. See Team Activity.

Organization totals#

The Totals tab provides a high-level overview of AI activity across your organization. Use it to monitor request volume, token consumption, service reliability, and usage trends over the selected time period before investigating individual models, users, agents, or teams.

The tab includes two charts:

• Total Tokens displays token consumption over time, including input tokens, output tokens, total tokens, and cache read tokens. Use this chart to understand how AI resources are consumed, identify changes in prompt and response sizes, and evaluate the effectiveness of cache reads when caching is enabled.

• Total Requests displays request volume over time, including successful requests, failed requests, and total requests. Use this chart to monitor traffic patterns, detect request spikes, and identify whether increases in failed requests correlate with higher request volume.

Use the Totals tab as the starting point for analyzing AI usage. Once you identify unusual trends, continue in Cost, Model Activity, User Activity, Agent Activity, or Team Activity. For actionable scenarios, see Use Cases — Organization totals.

Cost breakdown#

The Cost tab helps you understand where AI spending comes from, how it changes over time, and which models and users contribute most to your organization's AI costs. Use this tab to analyze spending trends, identify cost drivers, and evaluate the financial impact of AI usage before investigating individual models, users, agents, or teams.

The tab includes three charts:

• Cost Breakdown displays AI spending over time. Use it to monitor cost trends, identify spending spikes, and determine whether changes in usage, model selection, or organizational policies affect overall AI costs.

• Models Breakdown shows how costs or token consumption are distributed across AI models. Use it to identify the most expensive models, determine whether premium models are being used appropriately, and evaluate whether alternative models could provide similar results at a lower cost.

• Users Breakdown displays AI costs or token consumption by user. Use it to identify power users, understand who generates the highest AI expenses, and investigate unusually high resource consumption.

Together, these charts help you distinguish between costs driven by expensive models and costs driven by high request volume or token consumption. By switching between Spend and Total tokens, you can determine whether increased spending is caused by premium model usage or higher AI activity.

For actionable scenarios—cost drivers, budgeting, chargeback, and optimization validation—see Use Cases — Cost breakdown.

Activity tabs#

The Model Activity, User Activity, Agent Activity, and Team Activity tabs provide scoped analytics for a selected model, user, agent, or team—usage, cost, token consumption, and operational health. Use them after Totals or Cost when you need to attribute trends to a specific entity.

• Model Activity — Selected AI model: usage, cost, tokens, and reliability.
• User Activity — Individual user: usage, cost, preferred models (Model Usage), and reliability.
• Agent Activity — AI application or agent: workload, model mix, cost, and reliability.
• Team Activity — Organizational team: departmental adoption, spend, and performance.

Summary cards and charts (for example Cost Breakdown and Total Tokens) update for the selected entity and time range. For actionable scenarios, see Use Cases — Activity tabs.

Model Activity#

The Model Activity tab provides usage, cost, token, and reliability analytics for a selected AI model. Select a model from the dropdown to filter summary cards and charts—such as Cost Breakdown and Total Tokens—for that endpoint.

For scenarios such as adoption, cost comparison, and anomaly investigation, see Use Cases — Model Activity.

User Activity#

The User Activity tab provides usage, cost, preferred-model, and reliability analytics for an individual user. Select a user from the dropdown to filter summary cards, charts, and the Model Usage section for that account.

For scenarios such as power-user analysis, governance, and unusual usage investigation, see Use Cases — User Activity.

Agent Activity#

The Agent Activity tab provides usage, cost, model distribution, and reliability analytics for an AI application or agent. Select an agent to filter summary cards, charts (for example Cost Breakdown and Total Tokens), and Model Usage for that application.

For scenarios such as agent cost tracking, model utilization, and application troubleshooting, see Use Cases — Agent Activity.

Team Activity#

The Team Activity tab provides usage, cost, model, and reliability analytics for an organizational team. Select a team to filter summary cards, charts, and Model Usage for departmental adoption, spend, and performance.

For scenarios such as budgeting, governance, and AI initiative impact, see Use Cases — Team Activity.