Skip to main content

CAPTCHA guide

Slider CAPTCHA

The generic drag-the-piece puzzle, where the drag path matters as much as the target.

Difficulty: Medium
Image puzzle
Exploded concept drawing of how the Slider CAPTCHA challenge is built

"Slider CAPTCHA" describes a whole family of challenges rather than one product. The usual parts are a background with a gap, a puzzle piece, and a track with a handle. The visitor drags the piece into the gap, or turns a disc, matches two shapes or lines up a slider. Sites like them because they are cheap to build, work well on phones and slot easily into a risk engine.

How to identifyLink to How to identify

  • A background with a cut-out and a separate piece
  • A track with a handle and a label like "Slide to verify"
  • Canvas or <img> elements for the bg and the piece
  • A rotate version with an inner disc inside an outer ring
  • A click version asking you to "select the matching shape"
  • Network responses delivering the challenge images and a session ID

Common variantsLink to Common variants

Variant What you do Main difficulty
Jigsaw slide Move the piece into the gap Finding the gap
Gap fill Drag a shape onto the hole that fits it Matching shapes
Rotate Turn the disc to the right angle Estimating the angle
Multi-slide Line up several sliders Several coordinates
Click-match Click two identical objects Detecting objects

How it worksLink to How it works

  1. The server creates a background, a piece and the correct offset, sometimes with a tolerance.
  2. The client draws the challenge and records the drag or the clicks.
  3. On submit it sends the offset, the angle or the click coordinates, often with the full path and timing attached.
  4. The server accepts answers within the tolerance and may also rate how human the path looks.
  5. A failure triggers a new challenge. A success returns a token or flags the session as verified.

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

Tool What it does Success Rate Link
OpenCV matchTemplate Locates the gap through template matching or edge correlation 80–95% opencv/opencv-python
PuzzleSolver A complete slider solver with natural drag paths 80–95% vsmutok/Puzzle-Captcha-Solver
ddddocr Gap detection and basic slider helpers 70–90% sml2h3/ddddocr
gtkit/slider A Go solver with edge and template finders plus trajectory models 75–92% gtkit/slider
chushiyan/slide_captcha_cracking An OpenCV walkthrough with a full example 70–90% chushiyan/slide_captcha_cracking
DrissionPage / Playwright Captures the images and drags in a real browser 80–95% g1879/DrissionPage · microsoft/playwright-python

Practical approachLink to Practical approach

  1. Take the original background and piece from the network response instead of a rescaled screenshot.
  2. Prepare them: grayscale, blur, then Canny or Sobel edges.
  3. Run matchTemplate with TM_CCOEFF_NORMED. The best match gives the top-left corner of the gap.
  4. Turn that position into a slider distance, accounting for the piece's starting offset and any CSS scaling.
  5. Drag along an eased path: speed up, overshoot a little, come back, and add slight vertical wobble.
  6. Wait for the success state. On failure, take a new challenge and use a different path.

Practical tipsLink to Practical tips

  • Always calculate with the images' natural size, since resizing throws the coordinates off.
  • Sometimes the gap is just a darker area rather than a jigsaw outline. Edge detection is a good fallback.
  • The most common failure is a straight drag at constant speed.
  • Some sliders only look at where you let go, others at the whole path. Plan for the strict kind.
  • Rotate variants need angle estimation, such as feature matching or a trained model, not matchTemplate.
  • Budget your retries: a wrong answer usually kills the challenge on the spot.

Skip the solver: let BrowserSolver handle Slider 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.