---
title: Microsoft Fabric costs
canonical: "https://cloudmonitor.ai/docs/using-cloudmonitor/reports/fabric/"
description: "What your Microsoft Fabric capacity costs, which workspaces and people are consuming it, and which workspaces have gone quiet."
---

Microsoft Fabric is billed as **capacity** — you buy a pool of compute and everyone's workspaces draw
from it. That makes it awkward to attribute: one invoice line, many teams.

This report unpicks it.

## The tabs

**Chargeback** — capacity usage by workspace, item, domain and user, priced against the bill. This is
the tab that turns one capacity charge into per-team numbers you can actually hand out.

**Capacities** — your Fabric capacity inventory: size, region, state, which workspaces are assigned
to each, and what it costs. Each opens in Fabric or Azure.

**Workspaces** — governance across the tenancy: what exists, when each was last active, who's likely
using it, and cleanup commands for the ones that aren't.

## Workspaces is where the easy money is

Fabric capacity is bought ahead of use, so an abandoned workspace doesn't produce a visible charge —
it consumes headroom, and you notice when you're asked to buy more.

Sorting by last activity and looking at the stale end is a five-minute job that regularly pays for
itself. A workspace nobody has opened in six months is a conversation, not necessarily a deletion.

## A tab you may not have

**Capacity Metrics** — utilization, throttling, overages and storage — is an optional module. If it
isn't in your tabs, it isn't switched on for your workspace; an admin can enable Fabric Monitoring
under Settings ▸ Workspace Settings ▸ Features.

Chargeback, Capacities and Workspaces are unaffected either way.

:::note[Some capacity telemetry is illustrative]
Capacity **cost** is your real spend. Some of the fine-grained consumption telemetry behind the
metrics surface is a worked example rather than your own usage, because it needs a feed that isn't
in place yet.

The distinction matters if you're about to charge a team: the money is real, some of the
attribution detail is not yet. Each panel says which it is.
:::

## Filters

Period and cost basis drive the capacity cost figures.

Provider is grayed out — Fabric capacity is Microsoft-only. Cost group and virtual tags don't apply
yet either: Fabric usage isn't allocated to cost groups, and its telemetry carries no resource tags.
The chips say so rather than doing nothing quietly.

That last one is worth planning around. If you want Fabric costs inside a team's cost group total
today, that attribution has to come from the Chargeback tab rather than from the cost group filter.

## The page's own toggle

The Chargeback tab has a control for how capacity usage is attributed. It's independent of the global
filters — it picks the attribution method, not the scope — so changing it changes how the same total
is divided, never the total itself.