Tap4 AI Crawler is an open source web crawler built by tap4.ai, that will convert the website into the website summarize info with LLM. Includes powerful scraping, crawling and data extraction capabilities, web page screenshots. With Tap4 AI Crawler, you can not only easily update the ai tool detail for your AI Tools Directory but also summary of the website.
This project is based on Python, very lightweight, easy to maintain, suitable for individual developers interested in AI tools directories, and also for learners interested in Python. We welcome everyone to fork and star.
English | 简体中文
- Fetching titles, descriptions, and introductions of input websites
- Making screenshots of the input websites
- Support for using LLM (llama3/chatgpt) to process website introductions and generate SEO Friendly Markdown descriptions
- Quick configuration
- Fast deployment
Please follow our Twitter: https://x.com/tap4ai
If you find the project helpful, please consider buying me a coffee:
If you are interested in the project, please add my WeChat: helloleo2023, note: "tap4 ai open source", or scan the QR
code: 
- Register on Cloudflare
- Select R2 Service and create a bucket for image store, set for public access(option: set custom domain) and edit CORS Policy.
- CORS Policy as below:
[
{
"AllowedOrigins": [
"*"
],
"AllowedMethods": [
"GET",
"POST",
"PUT",
"DELETE",
"HEAD"
],
"AllowedHeaders": [
"*"
]
}
]-
Create R2 API Tokens for R2 API and select permission with Object Read & Write. Save your own params: ENDPOINT_URL, BUCKET_NAME, ACCESS_KEY_ID, SECRET_ACCESS_KEY, CUSTOM_DOMAIN. The params will config in the .env for tap4-ai-crawler.
-
-
Create a new project and service on Zeabur Note: Web scraping requires certain server configurations. It is recommended to purchase paid services from Zeabur and prioritize selecting U.S. nodes.
-
Fork tap4-ai-crawler to your own github and update .env params with your own.
Deploying the fork github repository in Zeabur, and configuring environment variables in Zeabur or manually modifying the .env file in the code repository. The environment variables are as follows:
GROQ_API_KEY: Key for Groq, apply for it HereS3_ENDPOINT_URL: Endpoint for S3(Recommand Cloudflare R2), apply for R2S3_BUCKET_NAME: Bucket name for S3(such as Cloudflare R2)S3_ACCESS_KEY_ID: Access key ID for S3(such as Cloudflare R2)S3_SECRET_ACCESS_KEY: Secret access key for S3(such as Cloudflare R2)S3_CUSTOM_DOMAIN: Custom domain for S3(such as Cloudflare R2), if you have a custom domain, fill it in; otherwise, it can be left blank.AUTH_SECRET: Custom access key for Rest API.
- Python 3.x version
git clone https://github.com/6677-ai/tap4-ai-crawler.git- Endpoint
- Access Key Id
- Secret Access Key
- Bucket Name
- Modify the
.envfile in the root directory with the following content, example:
## LLM Configuration: Large model related configuration
GROQ_API_KEY=gsk_********
## Object Storage Configuration: Storage related configuration
S3_ENDPOINT_URL=https://*****.r2.cloudflarestorage.com
S3_BUCKET_NAME=tap4ai
S3_ACCESS_KEY_ID=****
S3_SECRET_ACCESS_KEY=****
S3_CUSTOM_DOMAIN=****
AUTH_SECRET=****Install Python dependencies
pip install -r requirements.txtRun
python main_api.pyAfter running, a RestAPI will be exposed, access URL suffix: /site/crawl
Use curl to verify the API with POST request. Request params:
- Format: Json format,
- params: url (such as: https://tap4.ai) Request as below:
curl -X POST -H "Content-Type: application/json" -H "Authorization: Bearer xxxxx" -d '{"url": "https://tap4.ai", "tags": [ "selected tags: ai-detector","chatbot","text-writing","image","code-it"]}' http://127.0.0.1:8040/site/crawlResponse Params:
- Format: Json format
- Params: data-description: Description of website
- Params: data-detail: Detail content of website
- Params: data-screenshot_data: Screenshot of website
- Params: data-screenshot_thumbnail_data: Screenshot thumbnail of website
- Params: data-title: Title of website
{
"code": 200,
"data": {
"description": "Tap4 AI Directory is a tool provides free AI Tools Directory. Get your favorite AI tools with Tap4 AI Directory, Tap4 AI Directory aims to collect all the AI tools and provide the best for users.",
"detail": "### What is Tap4 AI?\n\nTap4 AI is an AI-driven platform that provides access to a vast array of AI technologies for various needs, including ChatGPT, GPT-4o for text generation and image understanding, Dalle3 for image creation, and document analysis.\n\n### How to Use Tap4 AI\n\nEvery user can utilize GPT-4o for free up to 20 times a day on tap4.ai. Subscribing to the platform grants additional benefits and extended access beyond the free usage limits.\n\n### Features of Tap4 AI\n\n#### Can I Generate Images Using Tap4 AI?\n\nYes, with Dalle3's text-to-image generation capability, users can create images, sharing credits with GPT-4o for a seamless creative experience.\n\n#### How Many GPTs are Available on Tap4 AI?\n\nTap4.ai offers nearly 200,000 GPT models for a wide variety of applications in work, study, and everyday life. You can freely use these GPTs without the need for a ChatGPT Plus subscription.\n\n#### How Can I Maximize My Use of Tap4 AI's AI Services?\n\nBy leveraging the daily free uses of GPT-4o document reading, and Dalle's image generation, users can explore a vast range of AI-powered tools to support various tasks.\n\n#### Will My Information Be Used for Your Training Data?\n\nWe highly value user privacy, and your data will not be used for any training purposes. If needed, you can delete your account at any time, and all your data will be removed as well.\n\n#### When Would I Need a Tap4 AI Subscription?\n\nIf the 20 free GPT-4o conversations per day do not meet your needs and you heavily rely on GPT-4o, we invite you to subscribe to our affordable products.",
"languages": [],
"screenshot_data": "https://demo.tap4.cn/tools/2024/6/15/tap4-ai-1718447471.png",
"screenshot_thumbnail_data": "https://demo.tap4.cn/tools/2024/6/15/tap4-ai-thumbnail-1718447477.png",
"tags": ["code-it","text-writing"],
"title": "Get your best AI Tools | Tap4 AI Directory",
"url": "https://tap4.ai"
},
"msg": "success"
}- Due to potential anti-scraping measures on the website, crawling may fail, and manual secondary checks are required.
- The information processed by the LLM may not meet expectations; you can try optimizing the prompt content yourself.
- The content processed by the LLM may still be the template passed through the prompts, which is also caused by anti-scraping measures and requires manual secondary checks.
- Web scraping requires certain server configurations. It is recommended to purchase paid services from Zeabur and prioritize selecting U.S. nodes.
The Collection for the AI tools all over the world. | Collect free ChatGPT mirrors, alternatives, prompts, other AI tools, etc. For more, please visit: Tap4 AI
Here is the website list for submitting your product to get users. Please visit StartUp Your Product List
Free Stable Diffusion 3 Online
Tattoo AI Design is a tattoo AI generator and design tool for tattoo fans. If you are interested, visit Tattoo AI Design
