Alibaba · Answers
Qwen-User
The `-User` suffix is the only thing anyone has to go on here, and a suffix is not documentation. Other labs use that convention for a fetcher that fires once, on demand, when a person in a chat session asks about a page, but Alibaba names no agent anywhere, in any language, and reading a crawler's behaviour out of a naming convention is how directory entries turn into fiction. What Alibaba does publish is a Model Studio training-data disclosure enumerating four data categories, whose web entry reads only "Publicly available internet data — web content accessible to the general public", with nothing about how that content is obtained or under what identity. The silence is confirmed rather than assumed: `qwen.ai` serves no robots.txt at all, and `/bot`, `/bots`, `/crawler` and `/webmasters` fall through to the app rather than 404, which also leaves no published range, hostname scheme or request signature to check a claimed hit against.
- Operated by
- Alibaba
- Purpose
- Answers
- robots.txt token
- Qwen-User
- Verification
- User agent only
Fetches a page in real time to answer a question someone just asked.
How to verify Qwen-User
Alibaba publishes no list of IP addresses for Qwen-User, so nobody (us included) can prove that a request carrying this user agent really came from Alibaba. The name is trivial to copy. Treat it as a claim the visitor is making about itself, not as an identity anyone has checked.
There is no range list to confirm. This entry was last reviewed against Alibaba's own documentation on .
User agent
Alibaba has not published a full user-agent string for Qwen-User. Requests are identified by the Qwen-User product token appearing in the User-Agent header; we match that token rather than a whole string, because the rest of the header varies and matching it would miss real traffic.
robots.txt for Qwen-User
Block
User-agent: Qwen-User
Disallow: /Allow
User-agent: Qwen-User
Allow: /robots.txt is a request, not an enforcement mechanism. It is honoured by convention, and a crawler that ignores it is stopped at your edge, not in a text file.
What blocking Qwen-User costs you
The reflex framing (that you would be opting out of a China-only surface) does not survive a look at the product. The public shared-conversation index under `chat.qwen.ai` is dominated by Russian, Portuguese, Arabic, Turkish and Spanish thread titles, so a European or American publisher blocking this token is not declining a market it never served. What a block would actually forfeit, *if* the live-fetch behaviour is real, is the cited snippet shown to someone who asked a question your page answers, and since Alibaba documents no such mechanism, that loss is asserted by third parties rather than by the vendor, and we are not going to price it for you.
Vendor documentation
Alibaba does not publish documentation for Qwen-User that we could find. Everything on this page comes from what they do publish elsewhere and from observed behaviour, so treat it accordingly.