GeeTest solving, one API.
Slide, click and intelligent GeeTest challenges, cracked by neural solvers in a fraction of a second.
No solving farms, no browser fleets to babysit. Send the site key and URL to /createTask, poll for the token, drop it into your request.
01 · DETECT
Send the challenge
Site key, URL and type to /createTask.
02 · SOLVE
Neural solver
ML + fingerprint emulation crack it.
03 · SHIP
Get your token
Poll /getTaskResult, use the token.
Ship it in five lines.
# pip install capbypass from capbypass import Client client = Client(api_key="cb_live_...") token = client.solve( type="GeetestTask", websiteURL="https://target.com", captchaId="e392...b4c1", ) print(token) # solved in 430ms
// npm i @capbypass/sdk import { Client } from "@capbypass/sdk"; const client = new Client({ apiKey: "cb_live_..." }); const token = await client.solve({ type: "GeetestTask", websiteURL: "https://target.com", captchaId: "e392...b4c1", }); console.log(token); // solved in 430ms
// go get github.com/capbypass/go client := capbypass.New("cb_live_...") token, _ := client.Solve(capbypass.Task{ Type: "GeetestTask", WebsiteURL: "https://target.com", CaptchaID: "e392...b4c1", }) fmt.Println(token) // solved in 430ms
GeeTest pricing.
| Variant | Price / 1,000 | Mode | Status |
|---|---|---|---|
| GeeTest | $0.99 | Proxy · Proxyless | Live |
// Billed per successful solve - failed solves cost nothing.All pricing
Which GeeTest types do you handle?+
GeeTest v3 (slide) and all nine v4 types - slide, icon, svg_icon, nine, word, phrase, space, pencil and voice. The solver auto-detects which one is served.
How do v3 and v4 differ in the request?+
One task type handles both. Send gt + challenge for v3; send captchaId (without challenge) for v4. Routing is automatic.
Do I have to specify the challenge type?+
No. You never set a risk_type - the solver detects the challenge and dispatches to the right routine.
What do I get back?+
The validated GeeTest seccode payload, ready to submit to the protected site.