# llms.txt - AI Crawler Information ## Product Overview Word2Mail adds a microphone button to Gmail and Outlook. Speak naturally, and it reads the thread you're replying to and turns your words into a complete draft — greeting, paragraphs, and sign-off included. Faster than typing, and it still sounds like you took your time. ## Main Features - **Voice-to-Email Drafting**: Dictate in Gmail or Outlook; the extension records the audio and sends it to our backend, which forwards it directly to Google's Gemini model in a single multimodal call that turns your voice into a properly structured, ready-to-send draft — no separate transcription step. - **Thread-Aware Replies**: When replying, Word2Mail reads the subject, sender, and visible body of the open Gmail or Outlook thread so the generated reply fits the conversation. - **Feedback-Driven Quality**: Rate any generated draft (like/dislike); ratings feed automated prompt and model selection, and every five ratings earns a bonus generation. - **Quota System**: Free tier includes 10 email generations/month. Paid plans and one-time Generation Packs unlock higher limits. - **Cloud Sync**: Preferences (language, tone, auto-stop duration) and custom instructions sync across devices via Firestore. ## Target Audience - **Primary**: Busy professionals, executives, and knowledge workers who use Gmail or Outlook and value communication efficiency. - **Secondary**: Non-native English speakers seeking help with email composition and structure. - **Tertiary**: Users with accessibility needs who prefer voice input. ## Key Pages - **Landing Page** (`/`): Product overview, feature highlights, and CTA to install the browser extension. - **Pricing Page** (`/pricing`): Free and paid plan tiers, plus one-time Generation Packs. - **Dashboard** (`/dashboard`): Authenticated users manage their quota, subscription, and preferences. - **Help Page** (`/help`): FAQ and troubleshooting for common issues. - **Privacy Policy** (`/privacy`): Data handling and security practices. - **Terms of Service** (`/terms`): Legal terms and conditions. ## Technology Stack - **Frontend**: React 19 with TypeScript, Vite, Vike (SSR/SSG) - **Backend**: Python FastAPI, Firestore, Firebase Authentication - **Browser Extension**: Manifest V3, provider-agnostic mail-control-layer for Gmail and Outlook integration, MediaRecorder-based audio capture - **AI**: Google Gemini API for multimodal (audio + text) email generation ## Data Handling - **Voice Audio**: Recorded in the browser extension and sent to our backend, which forwards it directly to Google's Gemini model. Processed entirely in memory on our servers — never written to disk, a database, or a log file — and discarded the instant the draft is returned. No stored copy of your recording. - **Email Thread Content**: Read from the open Gmail or Outlook tab, used only to generate the requested draft, and not persisted on our servers. - **Generation Records**: Metadata only (timing, model, token counts, rating) — never the input audio, instruction text, or generated content — retained 12 months. - **Analytics**: Privacy-first, consent-gated, cookieless event tracking via Plausible — no per-user identifiers, no email content, no voice data. See `/privacy` for the full data-handling account. ## Company & Support - **Website**: https://word2mail.com - **Chrome Web Store**: https://chromewebstore.google.com/detail/voicetomail-gmail-plugin/cjpkplljchcbddbnoepofbcnilkabbdf - **Contact**: admin@word2mail.com - **Privacy Policy**: https://word2mail.com/privacy - **Terms**: https://word2mail.com/terms --- Generated: 2026-07-07. Last fact-checked against the codebase: 2026-08-06.