What is slow replying costing you?
Enter your volume and your staff time. The calculator shows what the AI model costs on a frontier API and on the 4B model on our hardware, and how many hours you get back.
Calculator
Net benefit per month
- Staff hours saved
- Value of time saved
- Model cost on a frontier API
- Model cost on our 4B model
- Model cost you avoid
- Front office plan fee
- Your current AI tool today
How we worked this out
- The AI drafts each reply and you approve it in about 30 seconds. This assumes no draft needs editing. Drafts you rewrite take longer, so your real saving will be lower. Hours saved are your current weekly hours (x 4.33 weeks a month) minus that approval time.
- Tokens per message: short 600 in and 150 out, medium 2,000 in and 400 out, long 5,000 in and 800 out.
- Frontier API means Claude Sonnet 5.5 at $2.00 in and $10.00 out per 1M tokens, the OpenRouter list price as of 11 November 2026. The 4B model on our hardware is $0.10 in and $0.50 out per 1M tokens. A larger open-weight model through OpenRouter is charged at the provider's cost plus 30%.
- The plan fee is Front office at $99 a month. Net benefit is the value of time saved minus that fee and the 4B model usage.
- ChatGPT Team is assumed at $30 per seat per month. That is our assumption, not a quote, so check OpenAI's pricing page for the current price. Pesos use 1 USD = 57.5 PHP.
- This is an estimate, not a quote. Growth and Private plans have different fees and a setup cost.
Get this estimate and a plan.
Send us your numbers and we reply by email with a recommended plan. To get an invite code as well, request an invite. You can also print or save this page as a PDF for your own copy.
Sent.
We will reply to with a recommended plan. Want a copy now? Use Print / save as PDF above.