New: monitor which AI crawlers actually visit your site with Sona Agent Analytics  |  See the platform →

Kimi-User

Moonshot AI's fetcher for the Kimi assistant - notable for long-context reading of entire documents.

Reference reviewed by Sona

The operator publishes little or no documentation for this token. Details here come from community crawler registries and observed behavior, not vendor confirmation.

OperatorMoonshot AI
PowersKimi assistant live page fetches
PurposeUser-triggered fetch
User-agent tokenKimi-User
Respects robots.txtMostly

Kimi-User fetches a page when a user of Kimi asks a question that requires it. Kimi is Moonshot AI's assistant and one of China's most-used AI chat products, so the population able to trigger a fetch is large even though the crawler itself is thinly documented.

What distinguishes Kimi from the other assistants here is its long-context design - it built its reputation on ingesting very long documents in a single pass. If your site publishes lengthy reports, documentation, or reference material, a Kimi-User fetch is more likely to consume the whole thing than to sample the opening paragraphs.

Documentation is limited. Moonshot publishes no crawler page, no IP feed, and no compliance statement; community registries list it as a user-triggered fetcher with unclear robots.txt compliance. Given that nearly every documented user-triggered fetcher at other vendors declines to honor robots.txt, the prudent assumption is that this one does not either.

The audience question is the decision. If Chinese-language or China-market visibility has value to you, Kimi is a meaningful surface. If it does not, this is a straightforward block - with the caveat that a robots.txt rule may not be what enforces it.

How Kimi-User behaves

  • Long-context model, so a single fetch may consume an entire long document rather than a sample.
  • User-triggered, so traffic arrives in bursts tied to conversations rather than on a crawl schedule.
  • No operator crawler documentation, IP feed, or compliance statement.
  • Community registries record robots.txt compliance as unclear.

Allow Kimi-User

Kimi's large Chinese-speaking user base gets answers about your site grounded in your actual content, including your long-form material.

User-agent: Kimi-User
Allow: /

Block Kimi-User

You want no programmatic access from Chinese AI assistants - enforce at the server, since compliance is unconfirmed.

User-agent: Kimi-User
Disallow: /

How to verify Kimi-User

No published verification method

Moonshot AI publishes no IP range feed and no reverse-DNS convention, so Kimi-User traffic cannot be authenticated. Blocking has to rely on user-agent matching at the edge. Traffic typically originates from Chinese cloud infrastructure, which some operators use as a coarse signal - but that is geography, not identity, and will catch unrelated traffic alongside it.

Check an IP against this bot

Commonly confused with Kimi-User

ChatGPT-User

The same user-triggered role at a Western vendor, with the difference that OpenAI documents its position on robots.txt while Moonshot documents nothing.

TongyiBot

Another Chinese-operator token, but Alibaba's is associated with training and assistant content collection rather than purely live user fetches.

Kimi-User FAQs

Does Kimi-User respect robots.txt?

Unconfirmed. Moonshot publishes no compliance statement and community registries list it as unclear. Since nearly every documented user-triggered fetcher elsewhere declines to honor robots.txt, assume this one does not either.

What is Kimi?

Moonshot AI's assistant, among the most widely used AI chat products in China, known particularly for long-context reading - handling very long documents in a single pass.

Why would a Kimi fetch read my whole page rather than part of it?

Long-context capability is Kimi's design emphasis. Where other assistants sample or truncate, Kimi is more likely to ingest an entire long report or documentation page in one go.

How do I block Kimi-User?

User-agent filtering at your server, CDN, or WAF. There is no published IP feed to scope the rule against, so matching on the token is the practical approach.

Can Kimi-User read your page right now?

Test any URL and see exactly what AI crawlers receive.

Check my site