A site-specific cleaner removes only the keys on its list, so an unknown parameter stays in place: a page may need it for an article, product, search, or other behavior. The exceptions are links a cleaner rebuilds from scratch (Amazon product pages and Google Maps coordinates), which lose their query entirely. Each cleaner applies to the hosts under the site name and their subdomains unless the row says otherwise.
Social and messaging
| Site | Removes | Keeps |
|---|---|---|
| X (Twitter) twitter.com, x.com | s, t, ref_src, ref_url, twclid, via, and other share, attribution, timeline, app, and ad keys. | lang, and any key not on its list. |
| Instagram instagram.com | igsh, igshid, igsi, stkn, ig_rid, and other share, attribution, feed, story, and ad keys. | img_index, media_id, story_media_id, reel_ids, and any key not on its list. |
| Facebook facebook.com, fb.com | fbclid, __cft__, __tn__, mibextid, other share, attribution, session, Messenger, and ad keys, and any key that starts with fb_ unless it is on the keep list. | id, story_fbid, fbid, set, type, comment_id, and any key not on its list. |
| Reddit reddit.com, redd.it | share_id, ref, ref_source, ref_campaign, rdt_cid, correlator, and other session, notification, ad, and award keys. An out.reddit.com link is replaced by the destination in its url parameter. | context, sort, t, q, after, before, limit, and any key not on its list. |
| TikTok tiktok.com, tiktokcdn.com, tiktokv.com | _r, _t, _d, u_code, tt_from, share_app_id, share_link_id, and other share, device, session, e-commerce, and ad keys. | lang, q, t, video_id, item_id, and any key not on its list. |
| LinkedIn linkedin.com, lnkd.in | trk, trkInfo, trkEmail, lipi, refId, trackingId, the other li_ keys on its list, midToken and midSig, UTM keys, and other session, job-alert, invitation, and ad keys. | keywords, q, start, location and every f_ search filter, and any key not on its list. |
| Threads threads.net, threads.com | igshid, xmt. | Any key not on its list. |
| Pinterest pinterest.com | e_t, e_t_s, e_t_cs, ouuid, pin_unauth. | Any key not on its list. |
| Snapchat snapchat.com | share_id. | Any key not on its list. |
| WhatsApp whatsapp.com | link_source, link_medium. | Any key not on its list. |
Video and music
| Site | Removes | Keeps |
|---|---|---|
| YouTube youtube.com, youtu.be, youtube-nocookie.com | si, is, pp, feature, app, attribution_link, referral and UTM keys, and other playback, session, recommendation, notification, and ad keys; the same policy applies on YouTube Music. | v, t, list, index, start, end, search_query, and any key not on its list. |
| Twitch twitch.tv | tt_medium, tt_content. | Any key not on its list. |
| Spotify spotify.com | si, dl_branch, dl_mobileapp. | context, uri, and any key not on its list. |
| Netflix netflix.com | trkid, tctx. | Any key not on its list. |
| Bilibili bilibili.com | vd_source, seid, share_source, copy_link. | from, and any key not on its list. |
News and publishing
| Site | Removes | Keeps |
|---|---|---|
| The Guardian theguardian.com and www.theguardian.com only | cmp, in any letter case. | INTCMP, and any key not on its list. |
| The New York Times nytimes.com and www.nytimes.com only | smid. | smtyp, and any key not on its list. |
| Substack substack.com | utm_source, utm_medium, utm_campaign, isFreemail, isFromEmail, r, ref, and other email-click, attribution, newsletter, and A/B keys. | publication_id, post_id, and any key not on its list. |
| Wikipedia wikipedia.org | wprov. | Any key not on its list. |
| Medium medium.com | source, sk. | Any key not on its list. |
Shopping
| Site | Removes | Keeps |
|---|---|---|
| Amazon amazon.com, amazon.co.uk, amazon.de, and other country domains | tag, linkCode, ref, ref_, qid, sr, sprefix, crid, pf_rd_ and pd_rd_ placement keys, and other share, session, A/B, and cart keys on search and category pages. A product link (/dp/, /gp/product/, or a bare product ID) is rebuilt as https://<host>/dp/<ASIN>, without its query or fragment; the host is kept for the store domain and its www. and smile. hosts, and any other Amazon subdomain becomes www.amazon.com. | keywords, k, node, page, sort on search and category pages, and any key not on its list. |
| eBay ebay.com, ebay.co.uk, ebay.de | mkevt, mkcid, mkrid, campid, customid, toolid, ssspo, sssrc, ssuid, widget_ver, media, _trkparms, _trksid. | var, selected, hash, epid, and any key not on its list. |
| AliExpress aliexpress.com | spm, aff_fcid, aff_fsk, aff_platform, aff_trace_key, algo_pvid, pdp_npi, utparam-url, and other affiliate, page-position, share, and app-source keys. | sku_id, currency, language, gatewayAdapt, and any key not on its list. |
| Google Store store.google.com | hl, selections. | Any key not on its list. |
Search and maps
| Site | Removes | Keeps |
|---|---|---|
| Google Search google.com, google.co.uk, google.de, and other country domains | ved, ei, usg, sa, source, sourceid, sxsrf, oq, aqs, gs_lcp, and other click, session, client, image, news, shopping, and ad keys on /search pages. A /url redirect is replaced by the destination in its url (or q) parameter. | q, tbm, tbs, start, num, hl, gl, safe, and any key not on its list. |
| Google Maps maps.google.com, and google.com/maps links | A link with an @latitude,longitude,zoom path segment is rebuilt as www.google.com/maps/@latitude,longitude,zoom, without other path segments, the query, or the fragment. | Only the coordinates and zoom in the path. |
| Bing bing.com | cvid. | q, and any key not on its list. |
| DuckDuckGo duckduckgo.com | t, atb. | q, and any key not on its list. |
Redirect links it unwraps
A wrapper link that only forwards to another page is replaced by its destination, read from the wrapper itself. Nothing is fetched: FixupXer decodes one layer, accepts only a valid web address, and then cleans the destination too. A wrapper without a readable destination is not unwrapped; the site-specific and general cleaners still remove their listed keys from it.
| Service | Wrapper | Destination read from |
|---|---|---|
| Facebook link shim | l.facebook.com/l.php, lm.facebook.com/l.php | The u parameter. |
| LinkedIn safety redirect | linkedin.com/safety/go | The url parameter. |
| YouTube redirect | youtube.com/redirect | The q parameter. |
| Google Ads click | googleadservices.com/pagead/aclk | The adurl parameter. |
| Reddit Mail | click.redditmail.com | The percent-encoded second path segment, for example /CL0/<encoded URL>/…. |
| Bluesky Go | go.bsky.app/redirect | The u parameter. |
| GeoRiot / Geniuslink | target.georiot.com/Proxy.ashx | The GR_URL parameter. |
| LinkSynergy / Rakuten | click.linksynergy.com/link | The murl parameter. |
| Google Search /url | google.com/url | The url parameter (or else q; the Google country domains under Search and maps are covered too). |
| Reddit outbound link | out.reddit.com | The url parameter (if it is missing, the link is not unwrapped). |
Any website
After a site-specific cleaner, or on its own when none matches, the general cleaner removes well-known tracking keys from every link. It matches keys in any letter case, and it also drops a #Echobox= fragment.
| Kind | Keys |
|---|---|
| Campaign tags | utm_source, utm_medium, utm_campaign, utm_term, utm_content, utm_id, utm_referrer, utm_source_platform, and the other utm_ keys on its list. |
| Click identifiers | fbclid, gclid, gclsrc, gbraid, wbraid, dclid, msclkid, twclid, ttclid, li_fat_id, rdt_cid, srsltid, yclid, epik, and others. |
| Email and marketing IDs | mc_cid, mc_eid, _hsenc, _hsmi, __hstc, mkt_tok, vero_id, oly_anon_id, oly_enc_id, and others. |
| Analytics keys | _ga, _gl, _openstat, __twitter_impression, guccounter, guce_referrer, and others. |
| Key prefixes | Any key that starts with pk_, mtm_, matomo_, piwik_, itm_, xtor, at_, wt_, and others. |
| Kept on purpose | ref, source, si, from, and other keys a site may need. A site-specific cleaner can still remove them, as the YouTube and Spotify cleaners remove si. |
For example, a link like example.org/read?article=42&utm_source=chat becomes example.org/read?article=42: the campaign tag goes, the article selector stays.
- Removed
utm_source - Kept
article
- Before
example.org/read?article=42&utm_source=chat- After
example.org/read?article=42
The article value remains; the recognized campaign tag is removed.
When a site is not covered
A site can use identifiers that no cleaner recognizes, and none can promise to find every site-specific one. Review the result before you use it. For a site you visit often, you can add a custom rule and preview it in Test Lab. The lists on this page are copied from the app’s source code; the reference list in the app repository is the current one.
Social-link conversion (X readers, chat previews, and other frontends) is a separate, optional step. How readers and previews work.