Fieldbook to AIThe plain-English guide to which AI to use

Time-savers

Clear your inbox in minutes

Every new email gets sorted by urgency, and the ones that need a reply come with a draft waiting for you to review.

Last updated .

Costs here use the prices AI companies charge software developers, so they show relative cost, not your monthly bill. See what ChatGPT, Claude and Gemini charge you.

Starter recipe, not yet tested end to end.

What it does for you

Reads each new email, labels it (needs reply, just FYI, newsletter, receipt, spam), flags how urgent it is, and writes a short draft reply in your voice. It never sends anything on its own.

Time saved

Removes the first read-through and the blank-page problem of replying. We have not measured the time saved yet, and you still review every draft.

What it costs to run

Model to start withCost for 1,000 emails
Claude Haiku 5.5$0.30 (about 30 cents)
GPT-6 Luna$0.30 (about 30 cents)

Emails are short and the job is simple, so the cheapest reliable tier is enough. Estimate assumes about 1,500 tokens in and 300 tokens out per item (a token is about three-quarters of a word), at list prices. Real usage will differ.

How to build it

You will need: Zapier or n8n (automation apps); Gmail or Outlook

  1. Set the automation to start when a new email arrives.
  2. Send the sender, subject and body to the AI model with the instructions below.
  3. Read the reply (a label, an urgency, and a draft) and apply the label.
  4. Save the draft. Do not auto-send; read each one first.

Instructions to give the AI

You sort email. Reply with: category (one of: needs-reply, fyi, newsletter, receipt, spam), urgency (low, medium or high), and draft (a short, polite reply in my voice, or nothing if no reply is needed). Email:
{{email}}