Key facts
Runs entirely in the browser with no account, backend, or data collection.
Supports time-based jobs and token-throughput jobs in the same form.
Uses dated public GPU price examples as presets, but every rate can be overwritten.
Estimated total
$0.00
Runtime
$0.00
Daily burn
$0.00
Cost per 1M tokens
$0.00
How to interpret the result
Treat this page as a planning worksheet. Static examples and transparent formulas are useful for narrowing options, but real procurement still needs source quotes, workload benchmarks, storage and egress estimates, and failure-mode checks.
For buyer workflow context, read Buyer Guide, How Compute Is Priced, and GPU Cloud Price Comparison.
FAQ
Does the GPU cost estimator use live provider prices?
No. It includes dated public examples as presets and lets you overwrite the hourly rate with the quote you are actually evaluating.
Why does utilization matter?
A rented GPU still costs money while idle. Utilization adjusts token-throughput estimates so low batching or queueing does not look artificially cheap.
What should I add after this estimate?
Add storage, egress, model server overhead, retries, support commitments, and any minimum billing periods before using the result as a buying budget.