VAPTCHA is a Chinese CAPTCHA service. It started with click challenges and now mixes click, slide and 3D tasks, mostly on Chinese sign-up and login pages. Current integrations load the V4 SDK from cdn4.vaptcha.com, while older ones still run the V3 script and API.
How to identifyLink to How to identify
- The script
https://cdn4.vaptcha.com/src/v4.js(V4) orhttps://v-cn.vaptcha.com/v3.js(V3) - A global
vaptcha(...)initialiser called withvidandvkey - A button reading "点击完成验证" ("Click to verify")
- A click task showing several icons and an instruction
- Slide or 3D tasks for riskier sessions
- Token fields like
vaptcha_tokenorvaptcha-token
Common challenge typesLink to Common challenge types
- Click: pick the icons matching the prompt, in the right order.
- Slide: move a piece into the gap.
- 3D: rotate or manipulate an object.
- Invisible: low-risk visitors pass with no interaction.
How it worksLink to How it works
- The page initialises VAPTCHA with the public
vidand the server-sidevkey. - The SDK loads its settings (
v4c.vaptcha.com/api/v1/configin V4) and draws the widget. - Once the visitor solves the task, the SDK returns a token.
- The site's backend checks the token with VAPTCHA's API, usually
v41.vaptcha.com/api/v1/verifyin V4. Confirm the exact host and parameters in the current docs. - Tokens expire fast, work once and are tied to their challenge session.
Free & open-source bypass methodsLink to Free & open-source bypass methods
| Tool | What it does | Success Rate | Link |
|---|---|---|---|
| DrissionPage / Playwright | Drives the widget and grabs the token afterwards | 70–90% | g1879/DrissionPage · microsoft/playwright-python |
| YOLOv8 | Detects what to click in icon tasks | 80–95% (trained) | ultralytics/ultralytics |
| CLIP | Matches icons to the prompt without training | 55–80% | openai/CLIP |
| OpenCV | Finds the slider gap with template matching | 65–85% | opencv/opencv-python |
| Patchright / Camoufox | Lowers automation signals for invisible and passive modes | 60–85% | Kaliiiiiiiiii-Vinyzu/patchright-python · daijro/camoufox |
Practical approachLink to Practical approach
- Open the page and wait until the VAPTCHA widget has initialised.
- If the check is invisible, the token appears on its own. Just wait.
- Click tasks: capture the image, detect or classify the icons, then click them in the requested order.
- Sliders: work out where the gap is and drag along a natural path.
- Pull the token from the hidden field or the SDK callback and submit immediately.
- Use the same session and IP for solving and for verification.
Practical tipsLink to Practical tips
- VAPTCHA issues both client and server tokens. Make sure you send the one the backend is waiting for.
- V3 and V4 are separate APIs. Never mix their scripts, tokens or verification endpoints.
- Order matters in click tasks: the right icons in the wrong sequence still fail.
- Tokens expire quickly, so don't browse around between solving and submitting.
- VAPTCHA reacts to browser fingerprints. A real browser is generally more dependable than raw HTTP.