osFoundry / Platform
Usage
Tokens, runs, dollars — every cost, by category, in a chart you can drill into.
What it is
Usage is osFoundry's cost and activity dashboard. Because osFoundry is bring-your-own-key, the models you run bill straight to your own provider accounts with no reseller markup — which also means the spend is scattered across every provider console you use. Usage pulls it back into one place: it records the tokens consumed, the runs executed, and the dollar cost of each, then rolls them up into a single chart you can read at a glance or drill into by category.
The view is built around three units that answer three different questions. Tokens tell you what the model actually processed. Runs tell you how much work happened — how many chats, agent runs, or jobs went through. Dollars tell you what it cost. Every one of those is broken down by category, so instead of a single opaque total you can see which part of your workspace — which model, provider, or kind of activity — is driving the number. Because osFoundry spans one place for chat, agents, apps, code, knowledge and data, the category breakdown is where a spike stops being a mystery: you click into the segment that grew and see what it was.
This matters most for people running many models from many providers at once. Local models cost compute rather than API dollars, cloud and self-hosted models bill differently, and a single agent run can touch several of them. Usage is the one surface that reconciles all of that into comparable tokens, runs, and dollars — so you can catch a runaway agent, decide whether a task belongs on a local model instead of a cloud one, or simply know what the month is going to cost before the provider invoice arrives. It's local-first like the rest of osFoundry, so the record of what you ran stays in your workspace.
Key capabilities
- Tracks three units side by side — tokens consumed, runs executed, and dollar cost — so you see both the work done and what it cost
- Breaks every total down by category (model, provider, and kind of activity) instead of showing one opaque number
- Drill into any chart segment to see which part of the workspace drove a spike
- Reconciles spend across local, cloud, and self-hosted models into one comparable view — useful because BYOK spreads billing across every provider console
- Local-first: your usage record stays in your workspace, with costs billed at your own provider rates and no reseller markup
Use cases
- Solo BYOK user: opens Usage at month-end to reconcile osFoundry activity against provider invoices before the bill lands
- Agent builder: spots a run category climbing faster than the rest and drills in to find the loop that's burning tokens
- Cost-conscious team lead: compares dollar cost by model to decide which tasks should move from a cloud model to a local one
Frequently asked questions
Does osFoundry add a markup to the usage costs it shows?
No. osFoundry is bring-your-own-key, so models bill directly to your own provider accounts at your own rates. Usage reports what those runs cost you; there's no reseller margin added on top.
What's the difference between tokens, runs, and dollars in the chart?
Tokens are what the model processed, runs are how many chats/agent runs/jobs executed, and dollars are the cost. They answer three different questions — the volume of text, the amount of work, and the price — and each is broken down by category so you can see what's driving the number.
Can I see which model or activity is responsible for a cost spike?
Yes. Every total is split by category, and the chart is drillable — click into the segment that grew and it shows the model, provider, or kind of activity behind it, rather than leaving you with one lump sum.