For the complete documentation index, see llms.txt. Every page on this site is also served as Markdown: append `.md` to any URL, or send `Accept: text/markdown`.
Skip to content
Attensira LogoAttensira
Book a demo

Bots

What is Amazonbot, and should you allow it?

Amazonbot is Amazon's general crawler: Amazon says it improves Amazon products and services and may be used to train Amazon AI models. Block it if you object to that training use, and allow Amzn-SearchBot to stay eligible for Amazon search experiences such as Alexa. The three Amazon tokens are independent, so one rule never covers all of them.

Start freeBook a demo

7 days free · card required

AI traffic · attensira.com30 days to 4 Oct 2026

0requests filed as Amazonbot · 30 days

of 143,601 AI crawler requests from 7 operators

AI training crawlers on the board, beside attensira.com's rule for each
AgentAI requests, 30 daysattensira.com robots.txt
Amazonbot (this page)0Allowed
meta-externalagent131,105Allowed
GPTBot≈830Allowed
ClaudeBot≈68Allowed
Applebot-Extended0Allowed
Google-Extended0Allowed
MistralAI-Training0Wildcard

3 of 7 had requests filed in the window. User agents as sent, not checked against address lists.

Analyst agentAnalyst · reads the AI traffic boardAI traffic board, read over the Attensira MCP, and our live robots.txt · · n=143,601 AI requests
Karl-Gustav KallasmaaKarl-Gustav Kallasmaa, Founder & CEOLast updated

How do I allow or block Amazonbot in robots.txt?

Paste this into the robots.txt at the root of your domain. Beside it, the group attensira.com serves Amazonbot today.

robots.txt · for your sitepaste-ready
# Keep Amazon's training-capable crawler out; keep the# search one in.User-agent: AmazonbotDisallow: / User-agent: Amzn-SearchBotAllow: /
Example · a group to paste, written for the token Amazonbot
attensira.com/robots.txtnamed · allowed
User-agent: AmazonbotAllow: /Disallow: /api/
As served at attensira.com/robots.txt ·

We name Amazonbot and allow it everywhere except /api/, the same as every AI and search crawler we allow by name.

Read the whole file, or write yours with the robots.txt generator.

Who runs Amazonbot, and what is it for?

Amazon runs it, and we file it under training. Amazon says it improves Amazon products and services and may be used to train Amazon AI models. Amazon's search experiences such as Alexa read through Amzn-SearchBot, a separate token, so a rule here does not decide that.

From Amazon’s documentation, read . Why a training fetch and a search fetch are different decisions: the AI crawler guide.

Can you keep out of Amazon's training and stay in Alexa?

Yes, because Amazon splits the decision across three tokens. Amazon describes Amazonbot as used to improve its products and services, helping provide more accurate information to customers, and says it may be used to train Amazon AI models1. The word that matters is may, and it is Amazon's. For a site owner, "may be used to train" and "is used to train" call for the same rule.

The other two tokens do other jobs, and Amazon states each setting is independent of the others6:

  • Amazonbot

    Amazon's documented job
    Improves products and services
    Training?
    May be used to train Amazon AI models

  • Amzn-SearchBot

    Amazon's documented job
    Search experiences such as Alexa
    Training?
    Amazon says it does not crawl for model training

  • Amzn-User

    Amazon's documented job
    Live fetches for a user's question
    Training?
    Amazon says it does not crawl for model training

The robots.txt block above disallows Amazonbot and allows Amzn-SearchBot: out of the training-capable crawl, still eligible for Amazon's search experiences. A robots.txt that names only Amazonbot treats Amazon as one crawler, which is how sites end up blocking all three tokens or none.

How do I verify that a visit really came from Amazonbot?

A user-agent header is free text, so a log line naming Amazonbot is a claim until you check it.

Match the visiting IP against the published ranges in the record.

Amazon publishes a separate address list per token, so a verified request tells you which of its three agents called, whatever header it sent. If a request claims to be Amazonbot and the address is on the Amzn-SearchBot list, believe the address.

Why didn't your robots.txt change take effect?

Because Amazon reads robots.txt on its own schedule. Amazon states its automated crawling respects the Robots Exclusion Protocol, honouring user-agent and allow/disallow directives3, and three details on the same page matter more than the directive:

  • Rules are read per host. shop.example.com needs its own file; the one at example.com does not govern it.
  • A cached copy up to 30 days old may be used, and setting changes may take around 24 hours to reflect4. Do not judge a rule on the day you wrote it.
  • If robots.txt cannot be fetched, the agents behave as if it does not exist3. A robots.txt that returns a 500 is an open door, not a closed one.

What controls does Amazonbot not support, and which does it?

It does not support crawl-delay5. A crawl-delay line for Amazonbot is a no-op that looks like a fix. If load is the problem, scope a Disallow to the heavy path, or rate-limit verified Amazon addresses.

What it does support is a per-page training control. Amazon states its agents respect rel=nofollow at link level and the page-level noarchive, noindex and none robots meta tags, and that noarchive means do not use the page for model training5. If your objection is to some pages rather than all of them, noarchive on those pages expresses it without touching robots.txt at all.

VerificationAmazon
IP range file
ip-addresses

Amazon publishes no reverse-DNS suffix for Amazonbot; verify by the range file.

Researcher agentResearcherTranscribed from Amazon documentation ·

How should you match Amazonbot in logs?

Amazon's example header is Amazonbot/0.12. Amazon writes the Chrome version as the placeholder W.X.Y.Z, so the string is a template, and 0.1 is Amazon's own version to increment. Match the bare token Amazonbot, case-insensitively, as a whole word. Keep it distinct from Amzn-SearchBot and Amzn-User, which share no prefix with it: a filter for "amazon" misses both of them.

How does Amazonbot compare with other training-capable crawlers?

Its shape is unusual. GPTBot is a crawler whose only documented job is training. CCBot builds an open archive others train on. Amazonbot is a general product crawler whose output may be used for training. All three are refused with a named robots.txt group, and none of those refusals costs you an AI citation, because a training fetch never ends in a link. The training-crawler comparison is in CCBot vs GPTBot.

What can Attensira show you about Amazonbot?

AI traffic, on Growth and Business, counts crawler fetches per operator and declared purpose at the edge, so after you split Amazon's tokens you can see whether the split held. It reports user agents as sent; checking Amazon's per-token address lists stays your step. See crawler logs and the robots.txt generator.

User-AgentAmazon
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amazonbot/0.1) Chrome/W.X.Y.Z Safari/537.36
Researcher agentResearcherTranscribed from Amazon documentation ·

Should I allow Amazonbot in my situation?

Concrete cases, and what each one implies for this agent.

  • You want to stay eligible for Amazon search experiences such as Alexa but not contribute to Amazon model training.

    Disallow Amazonbot, allow Amzn-SearchBot.

    Amazon documents the three tokens as independent settings, and states that Amzn-SearchBot does not crawl content for generative AI model training while Amazonbot's content may be used to train Amazon AI models. The split is the whole point of Amazon publishing three tokens instead of one.

  • You edited robots.txt yesterday and Amazonbot is still crawling the paths you disallowed.

    Wait out the cache window before assuming the rule is broken.

    Amazon states these agents fetch host-level robots.txt or use a cached copy from the last 30 days, and that changes may take around 24 hours to be reflected. A rule that appears ignored on day one is usually a cached file, not a violated directive.

  • Amazonbot is hitting a heavy endpoint and you reach for crawl-delay.

    Use a disallow rule for that path, or rate-limit at the edge. Do not rely on crawl-delay.

    Amazon states that these user agents do not support the crawl-delay directive. Writing one is not a partial mitigation, it is a no-op — and it looks like a fix in your config while the load continues.

What else do people ask?

Does blocking Amazonbot remove me from Alexa?
Not by itself. Amazon documents Amzn-SearchBot as the agent whose access makes your content eligible for search experiences such as Alexa, and states each of its three user-agent settings is independent. Block Amazonbot and allow Amzn-SearchBot to keep that eligibility.
Does Amazonbot support crawl-delay?
No. Amazon states its user agents do not support the crawl-delay directive. Writing one does nothing. If Amazonbot load is the problem, a path-level Disallow or a rate limit on verified Amazon addresses is the tool.
Why is Amazonbot still crawling a path I disallowed yesterday?
Amazon says its agents may use a cached copy of robots.txt from the last 30 days, and that setting changes may take around 24 hours to reflect. Give a new rule at least a day before judging it, and check that the robots.txt file actually loads.
Can I keep a page out of Amazon training without blocking Amazonbot?
Yes. Amazon states its agents respect the page-level noarchive robots meta tag, and that noarchive means do not use the page for model training. That gives you a per-page training control robots.txt cannot express.

Where do these facts come from?

7 claims on this page, each with the page it was read from; last checked . The numbers in the text point here.

  1. Amazon states that Amazonbot is used to improve its products and services, helping provide more accurate information to customers, and may be used to train Amazon AI models.

    developer.amazon.com · retrieved

    “Amazonbot is used to improve our products and services. This helps us provide more accurate information to customers and may be used to train Amazon AI models.”

  2. Amazon gives Amazonbot's example user-agent string as "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amazonbot/0.1) Chrome/W.X.Y.Z Safari/537.36".

    developer.amazon.com · retrieved · changes often, check the source

  3. Amazon states that automated crawling from its listed user agents respects the Robots Exclusion Protocol, honouring the user-agent and the allow/disallow directives, and that when a robots.txt file cannot be fetched the agents behave as if it does not exist.

    developer.amazon.com · retrieved

    “Automated crawling from these listed user agents respects the Robots Exclusion Protocol, honoring the user-agent and the allow/disallow directives.”

  4. Amazon states that these agents fetch host-level robots.txt files or use a cached copy from the last 30 days, and that user-agent setting changes may take around 24 hours for its systems to reflect.

    developer.amazon.com · retrieved

  5. Amazon states that these user agents do not support the crawl-delay directive, and that they respect rel=nofollow and the page-level noarchive, noindex and none robots meta tags, with noarchive meaning do not use the page for model training.

    developer.amazon.com · retrieved

    “They do not support the crawl-delay directive.”

  6. Amazon documents Amazonbot, Amzn-SearchBot and Amzn-User as three user agents whose settings are each independent of the others, and states that Amzn-SearchBot does not crawl content for generative AI model training.

    developer.amazon.com · retrieved

    “Each user agent setting is independent of the others”

  7. Amazon publishes Amazonbot's addresses at developer.amazon.com/amazonbot/ip-addresses/, separately from the lists for Amzn-SearchBot and Amzn-User.

    developer.amazon.com · retrieved

Want whatever it takes to win the answer?

Start freeBook a demo

7 days free · card required