Viewing your per-model call and inference-cost breakdown
Beyond top-line totals, the account-wide Usage view includes a breakdown of activity by AI model — so you can see which models you're actually using and what they cost.
What the per-model breakdown shows
The Usage view sums activity across all of your projects and, within that, provides a per-AI-model breakdown of:
- Calls made to each model
- Actual inference cost attributed to each model
This sits alongside the account-wide totals (projects, agent calls, CAD jobs, flow runs, and capacity-left percentage), giving you a model-by-model view of where your usage is concentrated.
How cost is expressed
The breakdown reflects actual inference cost. For context, on the managed-inference path credits map to real inference value — 1 credit equals $0.01 of actual inference value. Your overall balance is shown in both credits and USD in the balance and history view, so the per-model costs tie back to your real spend.
BYO and unlimited accounts
If you bring your own provider key (BYO), those runs are unmetered and show as free — they don't draw down managed capacity. And if your account is flagged unlimited, the Usage view shows "Unlimited" rather than finite figures.
Where to find it
Open /account, expand the Usage section, and look for the per-model breakdown beneath the account-wide totals. It's the best way to understand which models are driving your inference cost and to spot opportunities to switch your default model in Settings.