MailSentry reads every incoming email with an open-weight AI, distills it into a 3-line digest, and pushes it to your phone in seconds. The full email still lands in your inbox — you just stop drowning in it.
Get Early AccessBank alerts · Security warnings · 2FA codes · Order updates — in Chinese, at a glance.
Point your email at MailSentry via Cloudflare Email Routing, or let our hosted version poll your mailbox over IMAP. No app to install.
An open-weight LLM (Llama 3.2) extracts a one-sentence summary, the key facts, and anything you need to act on — in Chinese or English.
The digest lands on WeChat/WeCom as plain text. Glance, decide, move on. The original email is forwarded to your inbox untouched.
A newsletter email → three glanceable sections, generated in seconds.
No inbox opened. Nothing missed.
A 3B open model costs ~nothing per email. Closed-API per-token pricing would eat the entire margin.
The model is a one-line string. Run it on our Workers, or self-host with Ollama — your bank statements never have to leave your infrastructure.
A better open model drops? One-line diff. No vendor API changes, no surprise pricing.