使用 CaptchaAI 在 Django 应用程序中处理验证码

Django 遇到验证码通常两件事：给表单挂 Turnstile/reCAPTCHA 挡机器人，或者打通有验证码保护的外部站点。前者校验 token，后者主动识别换 token，思路相反。

场景一：给 Django 表单接入验证码校验

前端拿到的 token 必须在服务端二次校验，不然验证码形同虚设。

用 Turnstile 保护联系表单

隐藏字段接收响应值

forms.py 里加一个隐藏字段：

# forms.py
from django import forms

class ContactForm(forms.Form):
    name = forms.CharField(max_length=100)
    email = forms.EmailField()
    message = forms.CharField(widget=forms.Textarea)
    cf_turnstile_response = forms.CharField(
        widget=forms.HiddenInput(),
        required=True,
    )

服务端二次校验

表单校验通过后拿 cf_turnstile_response 问 Cloudflare 官方接口，success: true 才处理表单：

# views.py
import requests
from django.conf import settings
from django.shortcuts import render, redirect
from .forms import ContactForm

def contact_view(request):
    if request.method == "POST":
        form = ContactForm(request.POST)
        if form.is_valid():
            # Verify Turnstile token with Cloudflare
            token = form.cleaned_data["cf_turnstile_response"]
            verification = requests.post(
                "https://challenges.cloudflare.com/turnstile/v0/siteverify",
                data={
                    "secret": settings.TURNSTILE_SECRET_KEY,
                    "response": token,
                    "remoteip": request.META.get("REMOTE_ADDR"),
                },
            ).json()

            if verification.get("success"):
                # Process the form
                return redirect("success")
            else:
                form.add_error(None, "CAPTCHA verification failed")
    else:
        form = ContactForm()

    return render(request, "contact.html", {
        "form": form,
        "turnstile_sitekey": settings.TURNSTILE_SITE_KEY,
    })

模板挂前端脚本

模板里照常引入前端脚本和挂件：

<!-- templates/contact.html -->
<form method="post">
    {% csrf_token %}
    {{ form.as_p }}
    <div class="cf-turnstile" data-sitekey="{{ turnstile_sitekey }}"></div>
    <button type="submit">Send</button>
</form>
<script src="https://challenges.cloudflare.com/turnstile/v0/api.js" async defer></script>

场景二：用 CaptchaAI 识别外部站点的验证码

这才是 CaptchaAI 派上用场的地方——拿到受验证码保护的外部页面数据前，先识别验证码换 token，再带 token 请求目标接口。

动手前先确认自己有权访问目标数据。

封装 CaptchaAI 服务类

把提交、轮询封装成一个 service 类，视图、命令、Celery 都能复用：

# services/captcha_solver.py
import time
import requests
from django.conf import settings


class CaptchaSolverService:
    """Django service for solving CAPTCHAs via CaptchaAI."""

    API_BASE = "https://ocr.captchaai.com"

    def __init__(self):
        self.api_key = settings.CAPTCHAAI_API_KEY

    def solve_recaptcha_v2(self, sitekey, page_url, invisible=False):
        """Solve reCAPTCHA v2."""
        params = {
            "key": self.api_key,
            "method": "userrecaptcha",
            "googlekey": sitekey,
            "pageurl": page_url,
            "json": 1,
        }
        if invisible:
            params["invisible"] = 1
        return self._submit_and_poll(params)

    def solve_turnstile(self, sitekey, page_url, action=None):
        """Solve Cloudflare Turnstile."""
        params = {
            "key": self.api_key,
            "method": "turnstile",
            "sitekey": sitekey,
            "pageurl": page_url,
            "json": 1,
        }
        if action:
            params["action"] = action
        return self._submit_and_poll(params)

    def solve_image(self, image_base64):
        """Solve image/text CAPTCHA."""
        return self._submit_and_poll({
            "key": self.api_key,
            "method": "base64",
            "body": image_base64,
            "json": 1,
        })

    def get_balance(self):
        """Check API balance."""
        response = requests.get(f"{self.API_BASE}/res.php", params={
            "key": self.api_key,
            "action": "getbalance",
            "json": 1,
        }, timeout=30)
        return float(response.json().get("request", 0))

    def _submit_and_poll(self, params, timeout=120):
        """Submit task and poll for result."""
        # Submit
        response = requests.post(f"{self.API_BASE}/in.php", data=params, timeout=30)
        response.raise_for_status()
        data = response.json()

        if data.get("status") != 1:
            raise CaptchaSolveError(f"Submit failed: {data.get('request')}")

        task_id = data["request"]

        # Poll
        start = time.time()
        while time.time() - start < timeout:
            time.sleep(5)
            result = requests.get(f"{self.API_BASE}/res.php", params={
                "key": self.api_key,
                "action": "get",
                "id": task_id,
                "json": 1,
            }, timeout=30).json()

            if result.get("status") == 1:
                return result["request"]
            if result.get("request") == "ERROR_CAPTCHA_UNSOLVABLE":
                raise CaptchaSolveError("CAPTCHA unsolvable")

        raise CaptchaSolveError("Solve timed out")


class CaptchaSolveError(Exception):
    pass

在 settings.py 里配置密钥

Key 都放进 settings，不要写死在业务代码里：

# settings.py
CAPTCHAAI_API_KEY = "YOUR_API_KEY"
TURNSTILE_SITE_KEY = "0x4AAAAAAAC3DHQhMMQ_Rxrg"
TURNSTILE_SECRET_KEY = "0x4AAAAAAAC3DHQhYYY_secret"

常见故障速查（含后文异步/Celery 场景）

在视图和管理命令里调用识别服务

采集外部数据的视图

请求流程：

• 接收 target_url
• 识别 Turnstile 拿 token
• 带 token 请求目标接口

# views.py
from django.http import JsonResponse
from django.views.decorators.http import require_POST
from .services.captcha_solver import CaptchaSolverService, CaptchaSolveError

@require_POST
def scrape_external_data(request):
    """Solve CAPTCHA and fetch data from external CAPTCHA-protected site."""
    url = request.POST.get("target_url")
    if not url:
        return JsonResponse({"error": "target_url required"}, status=400)

    solver = CaptchaSolverService()

    try:
        # Solve the CAPTCHA
        token = solver.solve_turnstile(
            sitekey="0x4AAAAAAAC3DHQhMMQ_Rxrg",
            page_url=url,
        )

        # Use token to access the protected resource
        import requests as http_requests
        response = http_requests.post(url, data={
            "cf-turnstile-response": token,
        }, timeout=30)

        return JsonResponse({
            "status": "success",
            "data": response.text[:1000],
        })

    except CaptchaSolveError as e:
        return JsonResponse({"error": str(e)}, status=500)

写一个 Django 管理命令

跑批量任务时管理命令更顺手，需要三个参数：

• --type：recaptcha 或 turnstile
• --sitekey：目标页面的 sitekey
• --url：目标页面地址

# management/commands/solve_captcha.py
from django.core.management.base import BaseCommand
from myapp.services.captcha_solver import CaptchaSolverService


class Command(BaseCommand):
    help = "Solve a CAPTCHA and print the token"

    def add_arguments(self, parser):
        parser.add_argument("--type", choices=["recaptcha", "turnstile"], required=True)
        parser.add_argument("--sitekey", required=True)
        parser.add_argument("--url", required=True)

    def handle(self, *args, **options):
        solver = CaptchaSolverService()

        self.stdout.write(f"Solving {options['type']} for {options['url']}...")

        if options["type"] == "recaptcha":
            token = solver.solve_recaptcha_v2(options["sitekey"], options["url"])
        else:
            token = solver.solve_turnstile(options["sitekey"], options["url"])

        self.stdout.write(self.style.SUCCESS(f"Token: {token[:50]}..."))

        # Check balance
        balance = solver.get_balance()
        self.stdout.write(f"Remaining balance: ${balance:.2f}")

用法：

python manage.py solve_captcha --type turnstile --sitekey 0x4AAA... --url https://example.com

Django 4.1+ 异步视图接入 CaptchaAI

Django 4.1 起支持异步视图，把 requests 换成 aiohttp，提交和轮询都不卡事件循环。

国内装包慢，可换清华 TUNA 镜像源。

# views.py (async)
import aiohttp
import asyncio
from django.http import JsonResponse

CAPTCHAAI_API_KEY = "YOUR_API_KEY"

async def solve_captcha_async(request):
    """Async view for solving CAPTCHAs."""
    sitekey = request.GET.get("sitekey")
    page_url = request.GET.get("url")

    if not sitekey or not page_url:
        return JsonResponse({"error": "sitekey and url required"}, status=400)

    async with aiohttp.ClientSession() as session:
        # Submit
        async with session.post("https://ocr.captchaai.com/in.php", data={
            "key": CAPTCHAAI_API_KEY,
            "method": "turnstile",
            "sitekey": sitekey,
            "pageurl": page_url,
            "json": 1,
        }) as resp:
            data = await resp.json()

        if data.get("status") != 1:
            return JsonResponse({"error": data.get("request")}, status=500)

        task_id = data["request"]

        # Poll
        for _ in range(30):
            await asyncio.sleep(5)
            async with session.get("https://ocr.captchaai.com/res.php", params={
                "key": CAPTCHAAI_API_KEY,
                "action": "get",
                "id": task_id,
                "json": 1,
            }) as resp:
                result = await resp.json()

            if result.get("status") == 1:
                return JsonResponse({"token": result["request"]})

    return JsonResponse({"error": "timeout"}, status=504)

用 Celery 处理耗时的识别任务

有些识别请求要等好几秒到几十秒，交给 Celery 后台处理更合理。

定义 Celery 任务

# tasks.py
from celery import shared_task
from .services.captcha_solver import CaptchaSolverService, CaptchaSolveError

@shared_task(bind=True, max_retries=2, default_retry_delay=10)
def solve_captcha_task(self, captcha_type, sitekey, page_url):
    """Background CAPTCHA solving with Celery."""
    solver = CaptchaSolverService()

    try:
        if captcha_type == "recaptcha_v2":
            token = solver.solve_recaptcha_v2(sitekey, page_url)
        elif captcha_type == "turnstile":
            token = solver.solve_turnstile(sitekey, page_url)
        else:
            raise ValueError(f"Unknown type: {captcha_type}")

        return {"success": True, "token": token}

    except CaptchaSolveError as e:
        self.retry(exc=e)

在视图里触发任务

前端轮询状态即可：

# Usage in views
from .tasks import solve_captcha_task

def start_solve(request):
    result = solve_captcha_task.delay("turnstile", "0x4AAA...", "https://example.com")
    return JsonResponse({"task_id": result.id})

def check_solve(request, task_id):
    from celery.result import AsyncResult
    result = AsyncResult(task_id)
    if result.ready():
        return JsonResponse(result.get())
    return JsonResponse({"status": "pending"})

常见问题

Turnstile 一直提示 verification failed，要查什么？

先确认 secret key 和前端挂件的 site key 是不是同一对，多数时候是测试密钥混进了生产 settings。

CaptchaAI 识别应该同步调用还是放进 Celery？

面向用户的视图放进 Celery，避免等待 10 多秒；管理命令、后台脚本同步更简单。

一个 Django 项目能同时处理多种验证码吗？

可以，solve_recaptcha_v2、solve_turnstile、solve_image 是同一 service 的不同方法。

小结

Django 接入 CaptchaAI，核心是一个封装提交/轮询流程的 service 类：命令同步调用，异步视图换 aiohttp，耗时长的任务丢给 Celery。

相关文章

• GeeTest 与 Cloudflare Turnstile 对比
• Turnstile Token 校验后仍报 403 怎么排查
• Cloudflare Turnstile 挂件模式详解