> ## Documentation Index
> Fetch the complete documentation index at: https://docs.orq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Activity

> Track how the workspace uses LLMs: spend, requests, tokens, and caching across models, users, and harnesses.

**Activity** is the page **Orq.ai** opens on after sign-in. It reports LLM usage split across five tabs: **Overview**, **Explore**, **Spend**, **My activity**, and **MCP**.

Scope follows the selected project: with a project active, every tab reports that project, and with no project selected the tabs cover the whole workspace.

Open it from **Activity** in the top bar, or at `/home` in the workspace URL.

The time range picker and the refresh control sit in the page header and apply to every tab. The range defaults to the last 14 days and is written to the page URL, so a copied link opens the same window.

## Overview

The default view: four KPI tiles over the selected range, followed by the charts that explain them.

| Tile | Counts |
| - | - |
| **Total spend** | LLM cost in USD. |
| **Requests** | Requests the **AI Gateway** served. |
| **Token volume** | Input and output tokens, cached input included. |
| **Error rate** | Share of requests that failed. |

Below the tiles, **Usage by model**, **Request volume by model**, **Token breakdown**, **Usage type**, and **Prompt caching** each chart one cut of the same window. Every card carries an **Explore** link that opens that chart in the **Explore** tab.

![Activity Overview tab showing Total spend, Requests, Token volume and Error rate tiles, and a Usage by model bar chart.](https://mintcdn.com/orqai/yZjVsgj_u7hik_Ol/images/activity-overview-418.png?fit=max\&auto=format\&n=yZjVsgj_u7hik_Ol\&q=85\&s=bc0191e852e5e65b2174e22ec00885d8)

The Overview tab: KPI tiles over charts for the selected range.

## Explore

Build a chart from the metrics and dimensions in the query builder, then read the numbers behind it.

* **Query builder**: pick a metric (**Cost (total)**, requests, tokens, error rate, and the rest), a **by** dimension, and an optional subgroup. **Add** opens the **Top N ranking** and **Rollup** controls; **Clear all** resets the builder.
* **Chart**: switch between **Lines** and **Stacked bars**, or **Donut** and **Breakdown** when the **Rollup** is **total**, expand to full width, or select **Save to Monitors** to keep the chart as a [Monitor](https://docs.orq.ai/ai-studio/observability/monitors.md) widget.
* **Share of** card: the same metric as a share of the total, which sums to 100%.
* **Ranked table**: one row per value of the split, with requests, cost, average cost per request, in and out tokens per request, and total tokens.

![Activity Explore tab with the query builder set to Cost total by Model, a stacked bar chart across the range, a Share of Cost bar, and a ranked table listing five models.](https://mintcdn.com/orqai/yZjVsgj_u7hik_Ol/images/activity-explore-418.png?fit=max\&auto=format\&n=yZjVsgj_u7hik_Ol\&q=85\&s=3294aec3d89571bc0c31d894e088b568)

The Explore tab: the query builder above a chart, share card, and ranked table.

## Spend

Break spend down by dimension, then read the trend.

* **Breakdown dimensions**: **Projects**, **Providers**, **Models**, **API keys**, [**Identities**](https://docs.orq.ai/ai-studio/observability/identities.md), or **Harnesses**, each with share, spend, and tokens per row.
* **Spend by \<dimension>**: a donut of the same split.
* **Spend over time**: the same window as a line or bar chart.

![Activity Spend tab with the Models breakdown selected, a Models table showing share and spend for five models, a Spend by model donut, and a Spend over time chart.](https://mintcdn.com/orqai/yZjVsgj_u7hik_Ol/images/activity-spend-418.png?fit=max\&auto=format\&n=yZjVsgj_u7hik_Ol\&q=85\&s=4fdbfc4a3224edc245314386b1989a3f)

The Spend tab: a model breakdown, a donut, and spend over time.

## My activity

Requests made as the signed-in user, or with API keys that user owns, across every project the user can reach: requests per day, requests by hour of day, and the agents and clients behind them.

## MCP

Tool calls through [**MCP Gateways**](https://docs.orq.ai/ai-gateway/mcp-portal/mcp-gateways.md) in the current scope: calls over time, success rate, errors, p95 latency, and the most used servers and tools.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.