Idea Lab · Oct 23 2026 · Fleet winner · Market brief

Fleet Token Economics

The winning market brief promises "analysis of AI compute and cloud token economics" but includes no figures. This calculator does the analysis with your numbers: what the fleet's daily work costs on hosted APIs, what the same work costs on self-hosted VPS models, and where the two cross.

Workload

EXAMPLE DATA: every price below is a placeholder. Replace with your invoices.

Agent tasks per day Input tokens per task Output tokens per task Share a small local model can handle (%)

Hosted API prices (per million tokens)

Input $ Output $ Cached-input discount (%)

Self-hosted

VPS cost per month $ Tasks per day one VPS can serve Your upkeep hours per month Value of an hour $

Monthly cost

Claimed by the proposal, not verified

"Economic analysis confirmed": no analysis, prices or sources were included in the package.
Default inputs here are illustrative only. Check current provider pricing and your own logs before deciding anything.

What to try

Slide the local-model share to 0 and to 90% and watch the hybrid line move.
Raise upkeep hours: self-hosting is only cheap if your time is cheap.
Paste in last month's real token counts from each provider's usage page.