Token Cost Calculator
Paste a prompt or a document and see what it costs on each model, input and reply.
How it works
- 1
Paste the text you'll send: a prompt, a note, a whole document.
- 2
Set how long you expect the reply to be, in tokens.
- 3
Compare the input, reply and total cost across GPT, Claude and Gemini models.
Runs in your browser. Nothing is uploaded.
Keep writing in HermesMarkdown
A local-first Markdown editor that runs in your browser: notes stay on your device, with tables, diagrams and AI when you want it. No account needed.
FAQ
How is the cost calculated?
The text is tokenized in your browser, then the count is multiplied by each model's list price per million input tokens. The reply is priced the same way at the output rate, which is usually several times higher.
Are the counts exact for every model?
They use the o200k tokenizer, which is exact for the GPT-4o and GPT-5 families; OpenAI doesn't publish the tokenizer of its newest models, so treat those as very close rather than guaranteed. Claude and Gemini use their own tokenizers, so their rows (marked ≈) are estimates. Anthropic says Claude 4.7 and later produce about 30% more tokens than older models for the same text, so expect Claude's real count to run higher than the estimate.
Which prices does it use?
Each provider's standard list price per million tokens, as of the date shown under the table. Batch processing, prompt caching and long-context tiers change the real cost, so check the provider's pricing page before you budget.
Can I see what a note in my vault costs?
Yes. In the editor, choose More → Token Cost, or run Show Token Cost from the command palette, to price the note you have open.
Is my text uploaded?
No. Counting runs in your browser, in a background worker. Your text is never sent to a server, and it's kept only in this browser tab.