Rates
Know what goes into the estimate.
A simple starting point: input tokens, output tokens, and the model you choose. Explore our proposed rates before starting a pilot conversation.
Draft pilot rates
| Model | Input / 1M tokens | Output / 1M tokens | Status |
|---|---|---|---|
| Nemotron 3 Nano | $0.15 | $0.60 | Draft |
| Nemotron 3 Super | $0.25 | $1.25 | Draft |
| Nemotron 3 Ultra | $1.50 | $6.00 | Draft |
Price version draft-pilot-2026-09. Draft Rittenhouse rates in USD. Final prices will be confirmed before access begins.
The cost of a question
Make the numbers yours.
Try a sample workload, then adjust the token counts.
Estimated cost · draft rates
Input × $0.25 / 1M
Output × $1.25 / 1M
An estimate, not a quote. Final pricing will be confirmed before pilot access begins. This calculator makes no model calls.
Access and billing
Initial access is a controlled prepaid pilot or an agreed invoice. Paid API tokens are not issued from this website yet. Tell us about your project to discuss access.
Support during the pilot covers routing the inquiry and scoping the work. Hands-on evaluation or integration is a separate engagement.
Questions
How are reasoning tokens billed?
They may contribute to billable output according to the measured upstream accounting. They are not added twice if already included in completion totals.
What request size is being considered?
The proposed pilot limit is 32,768 total tokens, including up to 8,192 output tokens. Final limits depend on verification of the selected route.
What happens if a request fails?
Failure handling and settlement terms will be confirmed before paid access begins. This site does not accept payments or issue live API keys.
Are unused credits refundable?
Credit expiry, refunds, minimum top-up, and invoice eligibility are confirmed in the pilot conversation. They are not implied by this page.
Does this include scientific consulting?
No. Token rates cover API usage. Evaluation and integration work is quoted separately.