Skip to content

CAPTCHA Solving

CAPTCHAs are the most common blocker for browser automation. UseBrowser handles them in two ways: automatic solving via extensions, and agent-assisted human handoff.

Install a CAPTCHA-solving Chrome extension (like Omocaptcha or SadCaptcha) directly in UseBrowser. These extensions detect and solve CAPTCHAs automatically in the background — no agent intervention needed.

TypeMethod
hCaptchaToken injection via solver extension
reCAPTCHAToken injection via solver extension
Slider puzzles (Shopee etc.)Bezier-curve drag simulation

Set API keys in the extension’s settings. Once configured, most CAPTCHAs are solved transparently without interrupting the agent’s workflow.

Slider puzzles (common on Shopee, Taobao) require a different approach. The agent uses native drag simulation — a 3-phase Bezier-curve drag with realistic noise and velocity that closely mimics human hand movement.

Video
CAPTCHA solving — auto-solve and slider drag
Two scenarios back-to-back. First: a reCAPTCHA appears on a login page. The solver extension activates — the checkbox clicks itself, the challenge is solved via token injection, page continues loading. No agent intervention. Second: a Shopee slider CAPTCHA appears. The cursor moves to the slider handle, then drags it across with a natural Bezier-curve motion — slight acceleration, subtle wobble, smooth deceleration at the target. The puzzle solves. ~20 seconds.

For CAPTCHAs that can’t be auto-solved — or sites with particularly aggressive anti-bot detection — the agent falls back to human handoff. It pings you via Telegram, you solve the CAPTCHA in the remote viewer, and the agent continues.

This two-layer approach (extension auto-solve + human fallback) means automation rarely gets permanently stuck.