Home / Link cleanliness

The links we give, and the links they give

We do not sell placement and we do not add tracking parameters, so we measured where the click actually lands. For each of 10 tools we took one link from our catalogue and from four other directories, followed it with GET, and recorded the redirect count and the final domain.

Comparison (10 tools × 5 sites)

Tool	Ours	ai-bot.cn	aigc.cn	futurepedia	futuretools
51mee	Clean	Tracking params ?utm_source=ai-bot.cn	Not captured No result in site search	— not checked	— not checked
HeyFriday	Clean	Clean	Clean	— not checked	— not checked
ChatArt	Clean	Tracking params ?utm_medium=link&utm_source=aibot&utm_campaign=FYTChatArt&utm_term=chatart20260825	Not captured No result in site search	— not checked	— not checked
Flova	Clean	Tracking params ?utm_source=ai-bot.cn	Clean	— not checked	— not checked
创客贴	Clean	Tracking params aiart.chuangkit.com/matrix?utm_source=aibotQD	Tracking params ?utm_source=aigccn	— not checked	— not checked
讯飞星火	Clean	Tracking params ?ch=xfxh-aigjtf01	Clean	— not checked	— not checked
通义听悟	Clean	Tracking params ?utm_source=ai-bot.cn	Tracking params ?cid=41	— not checked	— not checked
文心一言	Clean	Tracking params yiyan.baidu.com/?utm_source=ai-bot.cn → lands on wenxin.baidu.com (1 hop)	Tracking params yiyan.baidu.com/?utm_source=aigcdh → lands on wenxin.baidu.com (1 hop)	— not checked	— not checked
ChatGPT	Clean	Tracking params ?utm_source=ai-bot.cn	Not captured No result in site search	Tracking params ?utm_source=futurepedia&utm_medium=marketplace&utm_campaign=futurepedia	Redirect /go/chatgpt-openai → chatgpt.com/chat?ref=FutureTools.io (2 hops)
Midjourney	Clean	Clean	Clean	Tracking params ?utm_source=futurepedia&utm_medium=marketplace&utm_campaign=futurepedia	Redirect /go/midjourney → midjourney.com/app/?ref=FutureTools.io (1 hop)

Clean = official domain, no query string, no redirect; Tracking params = official domain with tracking parameters; Redirect = goes through the directory's own redirect endpoint; Not captured = the site was not reachable / not listed (this does NOT mean the site is dirty); — not checked = not checked this round (Chinese tools were not checked on futurepedia/futuretools).

Criteria (reproducible)

Tracking parameters: the query string contains utm_*, f, ref, src, channel, ch, unit, spm, etc.
Redirect: the href domain is not the tool's own, or the path is a redirect endpoint such as /go/, /out/, /redirect/, /utms/; the landing page changes domain or appends parameters.
How we followed: GET only, following up to 10 redirects; we record the final URL's domain and the number of hops. We do not use HEAD — HEAD reports normal pages like heyfriday.cn / wenxin.baidu.com as 404/403.
Main-entry rule: one link per tool per site (the one the site shows by default); extra entries are noted, not mixed into the main rates.

Our own gap

Our own gap: 25 of 1,501 clean_url values (1.7%) still contain parameters; we are cleaning them. The 10 sampled here are 0.

Limits

Not reachable: toolify.ai, theresanaiforthat.com, topai.tools and aixploria.com all returned Cloudflare 403; aibase renders links with JS, so only 3 tool pages could be located. Not captured does not mean the site is dirty — it only means we could not read its href this round.

Browse all AI tools →