Skip to main content

CAPTCHA guide

Amazon CAPTCHA

Amazon's old-style 'type the characters' page, shown when browsing looks automated.

Difficulty: Low
Text & audio
Exploded concept drawing of how the Amazon CAPTCHA challenge is built

The classic Amazon CAPTCHA is the "Enter the characters you see below" page Amazon shows when its risk engine thinks a browser is automated. It is a distorted-text image, not the AWS WAF challenge: those are two separate products on different pages. OCR handles it easily, but passing it does not on its own repair a session Amazon has flagged.

How to identifyLink to How to identify

  • An "Amazon.com" title and the prompt "Enter the characters you see below"
  • A form posting to /errors/validateCaptcha with method GET
  • Hidden fields amzn and amzn-r
  • The answer box field-keywords
  • An image from images-na.ssl-images-amazon.com/captcha/...
  • A "Try different image" link and a "Continue shopping" button
  • The footer line "Sorry, we just need to make sure you're not a robot"

How it worksLink to How it works

  1. When a session looks suspicious, Amazon's risk engine redirects it to this page.
  2. The page carries hidden tokens (amzn, amzn-r) alongside the text image.
  3. The visitor types the characters into field-keywords and sends the form.
  4. Amazon checks the answer and the hidden tokens, then redirects back to the original path.
  5. Case usually doesn't matter, but every wrong answer pushes the risk score up.
  6. Passing resets the session's risk score for the moment. It does not earn lasting trust.

Amazon CAPTCHA vs AWS WAFLink to Amazon CAPTCHA vs AWS WAF

  • Amazon CAPTCHA: the old-style text image on amazon.com/errors/validateCaptcha.
  • AWS WAF CAPTCHA: a different AWS product with its own widget, puzzle and token. See Amazon WAF.

Free & open-source bypass methodsLink to Free & open-source bypass methods

Tool What it does Success Rate Link
ddddocr Reads Amazon's fairly clean text images very well 90–98% sml2h3/ddddocr
Tesseract + preprocessing Thresholding, deskewing and an alphanumeric whitelist 70–90% tesseract-ocr/tesseract
EasyOCR A sturdy default when images are noisy 80–95% JaidedAI/EasyOCR
Playwright / DrissionPage Keeps the session, sends the form and follows the redirect 90–98% microsoft/playwright-python · g1879/DrissionPage

Practical approachLink to Practical approach

  1. Stay in the browser session that got the CAPTCHA. Don't open a new one.
  2. Download the image and clean it: grayscale, more contrast, threshold.
  3. Read it with ddddocr or Tesseract, removing spaces and punctuation.
  4. Fill in field-keywords and submit with the original amzn and amzn-r values.
  5. Follow the redirect and confirm the product page actually loads.
  6. If the CAPTCHA keeps coming back, slow down and switch to a cleaner IP rather than retrying fast.

Practical tipsLink to Practical tips

  • Amazon generally ignores case, but submit the characters exactly as OCR read them.
  • The hidden amzn and amzn-r fields link your answer to the original request. Keep them intact.
  • A solved CAPTCHA won't make up for a poor IP or a robotic request rhythm.
  • For everyday browsing, skip incognito and fresh profiles. A signed-in account earns more trust.
  • Mobile pages lay the same CAPTCHA out differently. Recognise it by the form action rather than its look.
  • Don't mistake this page for the AWS WAF challenge, which needs a different approach.

Skip the solver: let BrowserSolver handle Amazon CAPTCHA

Website AntiBot runs a real browser session with matching TLS, headers, cookies and IP, clears the challenge, and returns the page. You send a URL, you get HTML back.

For education and authorized testing only. Respect the law, privacy, and the terms of the sites you work with. Difficulty labels are editorial estimates. Adapted from the CC0 community field guide on GitHub.