One prompt, every model. Paste the exact prompt you plan to ship to production, select your candidate models, and get an instant side-by-side cost ranking table you can share with your engineering team.
No prompt — preset volume:
Knowledge Base
Frequently Asked Questions
How does the prompt cost calculator work?
Paste your prompt, set the expected output size, and select the models you're considering. The calculator estimates tokens from the prompt text, then prices the request against each model's current published input and output rates, sorted cheapest first.
Can I share the results?
Yes — the copy button produces a plain-text ranking you can paste into Slack, a doc or a pull request, e.g. 'GPT-5 mini: $0.006 · Gemini 2.5 Flash: $0.0065 · …'.
Why is the cheapest model not always the right choice?
Price per token only covers cost. Capability, latency, context window and output-quality requirements decide whether a cheaper tier actually works for your task — the comparison table links to each model's full economics page.