Replying to email is mostly deciding on a tone, then finding the words. This tool does both — it reads the email you paste and returns three reply options at once, one formal, one neutral, and one friendly, each with its own subject line, using your own OpenAI or Anthropic key entirely in your browser.
How it works
Choose a provider and model, paste your API key, drop in the incoming email, and optionally add your context (your role, relevant history) and the outcome you want. The tool builds a prompt that requests exactly three replies across the three tones and asks the model to return strict JSON — tone, subject, and body per reply. The response is parsed and shape-checked in the browser before display. It is one direct request to the provider.
For Anthropic, the request includes the official direct-browser-access header so it works straight from the page.
Why three tones instead of one
Tone is the hardest thing to calibrate in email because it depends on the relationship, the stakes, and the context — and getting it wrong costs more than the words themselves. A reply that is too formal with a long-time colleague reads as cold or passive-aggressive; a reply that is too casual with a new external contact reads as unprofessional.
Getting three drafts side by side lets you:
- Compare and choose the register that fits the relationship and the moment
- Mix elements across tones — take the structure of the neutral draft and the sign-off of the friendly one
- Spot over-explanation — AI replies tend to over-explain; seeing three versions helps you see where to cut
- Confirm the right outcome — sometimes seeing the formal version of “decline this meeting” written out clearly confirms you want to decline; seeing the friendly version confirms you want to say yes with warmth
What each tone produces
Formal — professional register, complete sentences, no contractions, careful acknowledgement of the subject. Use for: new external contacts, senior stakeholders, complaints, legal or financial matters, first contact with a client or partner.
Neutral — clear and direct without being cold. Contractions are fine, structure is maintained, but the tone is collegial rather than corporate. Use for: internal colleagues, ongoing professional relationships, most routine business email.
Friendly — warm opener, more conversational phrasing, slightly more personal. Use for: people you know well, creative or informal contexts, internal team communication where a warmer tone fits the culture.
Steering the output
- Your context — “I am the project lead and the deadline already moved once.”
- Desired outcome — “Politely decline but keep the door open,” or “Confirm and propose Tuesday.”
The more you give the model, the less generic the drafts. Without context, the model treats the incoming email as its only information; with context, the draft reflects your actual position and relationship.
Tips
- Copy a draft, then trim it — AI replies tend to over-explain; one good cut sharpens them.
- Mix tones: take the structure of the neutral reply and the warm sign-off of the friendly one.
- Always proof anything that commits you to a date, price, or promise before hitting send.
- Your API key stays in your browser tab and is cleared on refresh — it never touches any Gera server.