ClearFlow

Hides spam, abuse and sexual replies on X, Reddit, YouTube, RedNote and Hacker News, showing why. Translates foreign posts.

ClearFlow makes comment sections readable again on X, Reddit, YouTube, Xiaohongshu (RedNote) and Hacker News. • Cleans up junk replies — spam and lead-generation bots, abuse and sexual content — and tells you why each one was hidden. Click to see it anyway; mark it when it got one wrong. • Hides content-free replies like "nice", "+1" or "following", with one line at the top of the comments to bring them back. • Merges runs of hidden replies into a single line; when a thread is full of junk, tucks it all away behind one summary line. • Learns from what you mark: hide a spam reply yourself and replies of the same kind — reworded, from other accounts, pushing the same scheme — are hidden too. Undo anytime. • Translates foreign-language posts and comments into your language, in place, with a switch back to the original. On X it replaces X's automatic translation only when the new one is clearly better. • Turn it on or off per site, and choose cleaning and translation separately. ClearFlow uses two services you sign up for and pay by usage, with your own API keys: TypeSafe (judging replies) and OpenRouter (translation). A setup page helps you enter them; the settings show what you've spent. Without keys, only the built-in local rules run. Privacy: the comments and posts you're viewing on the supported sites are sent to those two services to be judged or translated (with your most recently marked spam replies as examples, if you've marked any). Nothing is sent to us, and there is no analytics or tracking. Keys and history stay in your browser. Open source (MIT): https://github.com/remember2015/clearflow ClearFlow is not affiliated with X, Reddit, YouTube, Xiaohongshu or Y Combinator.