CAPTCHA / GEETEST_NINE

GeeTest / Nine

A captcha with a 3×3 grid: pick the cells that match the hint image. The hint is an object or an emotion (a crying, angry or smiling face); the server detects the task kind itself. The solver returns the cell indices.

Input

Prepare grid.jpg — the full 3×3 grid, and hint.png — the hint picture from the task. Do not cut out the nine cells yourself: the server splits grid.jpg.

0 1 2
3 4 5
6 7 8

Parameters

Parameter	Type	Required	Default	Description
image	string	Yes	—	Source image in base64.
ques	string[]	Yes	—	Non-empty array of base64 hints. This solver uses the first element: normally send exactly one hint.
prompt_key	string	No	Empty string	Original name or path of the hint image from the provider's data. The last path segment is used for the dictionary of known tasks. If the key is missing or not found, the hint is classified from the image.
picks	integer	No	3	How many cells to select. Usually 3. Use the number from the task.

Response

Recognition data	Type	Description
extra.cells	integer[]	Selected cells, indices 0–8: left to right, top to bottom. Sorted by index.
extra.debug.kind	string	object — the hint is an object; emotion — the hint is an emotion.
extra.debug.label	string	The object or emotion (crying, angry, smiling) that was searched for.
extra.debug.label_source	string	dict — the hint was found by prompt_key; image — the emotion was recognized from the hint image; zero-shot — the object was determined from the image.
extra.debug.sim	number[9]	Diagnostic match scores of the nine cells against the hint.

Applying the answer

Read result["extra"]["cells"]. For index i: column = i % 3, row = i // 3. Cell centre relative to a grid of width W and height H: x = (column + 0.5) × W / 3; y = (row + 0.5) × H / 3. Add the grid's position on the page and submit the verification.

Common mistakes

Numbering starts at 0: top-left square is 0, bottom-right is 8.
The grid and hint must belong to the same round.

Scenario errors and retries →

Request example

import base64
from pathlib import Path
import os
import uuid
import requests

def image64(filename):
    return base64.b64encode(Path(filename).read_bytes()).decode("ascii")

payload = {
    "image": image64("grid.jpg"),
    "ques": [image64("hint.png")],
    "picks": 3,
}

# One ID per task; keep it for network retries.
request_id = str(uuid.uuid4())
response = requests.post(
    "https://capi.z3n.pro/v1/solve/geetest_nine",
    headers={
        "Authorization": f"Bearer {os.environ['CAPI_API_KEY']}",
        "Content-Type": "application/json",
        "Idempotency-Key": request_id,
    },
    json=payload,
    timeout=90,
)
response.raise_for_status()
result = response.json()
for index in result["extra"]["cells"]:
    row, column = divmod(index, 3)
    print(index, row, column)

Payment and all routes →