Skip to content
botfilter

Announcement

Introducing Botfilter

Botfilter flags AI generated text on websites you visit. Filter AI slop in your feeds, AI spam in your inbox, AI generated articles, and more.

Drafted with AI from the source code and reviewed by Claude Opus 5.5 (claude-opus-5-5), an AI model.

A 38-second film with captions and no narration. AI-written text turns up in feeds, inboxes and articles; Botfilter outlines a likely AI-written post on an X-style timeline, folds one on a LinkedIn-style feed, keeps flagged email in the inbox list, marks one paragraph on a news page, shows the Fewer, Balanced and More settings, explains that the model runs on your device with no account, and shows that each mark is a probability you can correct with Not AI.

Botfilter flags AI generated text on websites you visit

Botfilter is a free browser extension that marks likely AI-written posts and comments on X, Reddit, LinkedIn, YouTube, and Hacker News, email in Gmail, and paragraphs elsewhere. Outline, Collapse, or Hide them.

Outline marks a post and leaves it where it is

On a social feed, Outline draws an orange box around each post that is likely AI-written. Hover it to see the probability, and click Not AI when the mark is wrong.

Collapse folds a post to one line

On a work feed, Collapse folds each flagged post to a single line. You keep your place in the feed, and one click brings the text back.

In Gmail, flagged email stays in your inbox list

In Gmail, Botfilter never removes an email from your inbox list. Even in Hide mode, a flagged message stays in the list, dimmed, with an orange stripe. If the opened message is also flagged, Hide hides its text.

On other sites, Botfilter checks each paragraph

On news sites, blogs, and other pages, Botfilter checks each paragraph, list item, and quote on its own, and skips any under 25 words. It can mark one passage and leave the rest of the page alone.

The model runs on your device

You don't need an account. Botfilter scores text with a 23.6 MB model inside your browser, so the text of the pages you read stays on your device.

More on this

It is for skimming feeds, not for judging one document

Botfilter is for people who skim feeds and inboxes and want to know what is likely AI-written. Pick Fewer, Balanced, or More marks. For a verdict on one essay or report, use a checker that publishes its error rates.

More on this

We plan to list Botfilter on Firefox Add-ons

We plan to list the Firefox build on Firefox Add-ons and keep publishing the test results for each new model.

Each mark is a probability, not a verdict

Marks can be wrong. Balanced aims to flag about 1 in 100 human posts. The highest rate for one source and length was 1.95%, on short web passages. Human posts on X, LinkedIn, YouTube, and Hacker News, and second-language English, are not measured.

More on this

Botfilter is not in a browser store yet

Status: In development. The Botfilter home page will link to its Chrome Web Store listing when it goes live.

Go deeper