Presend
The send-time gate for email
About Presend
Presend is a browser extension that sits in your email compose window and acts as a gate rather than an assistant. Where most email tools try to write for you, Presend does the opposite: it stands between you and the send button and stops the messages you would regret - a leaked API key, a reply-all to the wrong list, a draft that reads like ChatGPT wrote it, an off-voice tone with an important client. It works silently inside Gmail and Outlook, runs a set of pre-send checks the instant you are about to send, and for real risks it blocks the send until you confirm rather than showing a banner you will learn to dismiss. The pitch is that the average professional sends around 40 emails a day and each one is a small reputation decision, so the value is in catching the rare bad one before it leaves.
A core design choice is privacy. The first layer of rules runs locally in your browser, so leaked-secret detection, recipient verification, broken-link checks and AI-slop scoring cost nothing and add no latency. Because it does not use Gmail API scopes, there is no restricted-scope OAuth and your text does not leave the browser without your own key. For the AI-powered checks Presend uses a bring-your-own-key model, so you connect your own provider key and your content is processed on your terms. That combination - local-first rules plus BYOK for anything heavier - is what lets it market itself as a send-time safety net you can trust with sensitive mail rather than another cloud assistant that ingests your inbox.
You add Presend to Chrome in one click and it starts working inside Gmail and Outlook with no new app to learn. You compose as you always would, and Presend watches in the background, running ten or more checks the moment you are about to send. When it finishes it shows a voice score, an AI-slop score and any findings - leaked secrets, unverified or external recipients, broken links, or phrasing that reads as AI-generated. Low-risk issues are flagged for you to review; genuine risks trip the send-guard, which blocks the send until you explicitly confirm. If the draft reads like slop, a one-click rewrite strips the AI phrasing while keeping your voice, with options like More Me, Shorter and Less AI. You can also write your own checks in plain English, and those run only on your mail at send time.
- •Pre-Send Safety Net - Local rules catch leaked secrets and PII, wrong recipients, broken links and AI-slop the moment before you hit send, with zero AI cost and zero latency
- •Send-Guard - Real risks do not just get flagged, the send is blocked until you confirm, so it behaves as a gate instead of a warning you dismiss by day two
- •Voice And Slop Scoring - An AI score from 0 to 100 rates authentic voice and AI-generated content, so you catch the drafts that read like a machine wrote them
- •Voice-Preserving Rewrite - One click strips AI-slop while keeping your own voice, so the email stops sounding like ChatGPT rather than more like it
- •Your Own Checks - Describe any rule in plain English, such as do not let me sound passive-aggressive or always include a next step, and it runs at send time on your mail only
- •Custom Actions - Build multi-step workflows with webhooks, conditions, MCP connectors and AI prompts, similar to GitHub Actions but for email policy
- •Local And Private - Layer-one rules run in your browser with no Gmail API scopes, so there is no restricted-scope OAuth and your text never leaves without your key
Presend is built for anyone whose outbound email carries real consequences. Salespeople and account executives who fire off dozens of messages a day get a backstop against the reply-all disaster and the tone that lands wrong with a prospect. Engineers and technical staff get protection against the single most expensive mistake in this space - pasting a live API key or credential into a message to a client. Customer success, HR and operations teams get the external-recipient warning that stops sensitive data going to the wrong inbox. For teams, the Organisation tier turns individual habits into enforced policy: org-wide custom rules, mandatory disclaimers, no-PII rules and a shared analytics view of risky sends caught over time. It suits people who explicitly do not want an AI to write their emails and only want the bad ones stopped before they go out.
Pricing
$0 - $9/mo
- Free$0/mo
- Pro$3/mo
- Organisation$9/mo
From the vendor pricing page, 2026-09-27
