No need to learn CSS selectors or XPath — just describe what you want in plain English and AI Parser understands.
Fully control output format with JSON Schema, ensuring consistent and predictable data structures every time.
Leverage the most powerful AI models including GPT-4o and Claude for high-accuracy data extraction.
Get started immediately with no credit card required — the free tier is perfect for small projects.
curl -X POST https://api.proxyturk.com/v1/parse \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"url": "https://example.com/products",
"prompt": "Extract all product names, prices, and stock availability from this page",
"format": "json"
}'{
"success": true,
"data": [
{
"name": "Wireless Bluetooth Headphones",
"price": "$49.99",
"in_stock": true
},
{
"name": "USB-C Hub Adapter",
"price": "$29.99",
"in_stock": true
},
{
"name": "Mechanical RGB Keyboard",
"price": "$89.99",
"in_stock": false
}
],
"credits_used": 1
}Choose the plan that fits your project. All plans include proxy rotation, anti-bot bypass, and JavaScript rendering.
Custom volume, SLA guarantee, 1,000+ req/min rate limit, and dedicated integration support for enterprise needs.
| Endpoint | Credits | Description |
|---|---|---|
| AI Parse | 1–3 | Base 1 credit, +1–2 for high token usage |
| SERP API | 3 | Per search query |
| Web Scrape (basic) | 2 | HTML/Markdown output |
| Web Scrape (JS) | 5 | With JavaScript rendering |
| Web Scrape (screenshot) | 7 | With screenshot capture |
| Web Scrape (actions) | 10–12 | With browser actions |
AI Parser is a revolutionary tool that uses artificial intelligence models (GPT-4o, Claude) to extract structured data from any HTML page. Unlike traditional scraping methods, you don't need to write CSS selectors or XPath expressions. Define the data you want with natural language commands, and AI Parser returns clean, organized JSON output.
Use it alongside our Web Scraping API to fully automate your data extraction pipeline. For search engine result pages, integrate our SERP API service, and for general web scraping operations, combine our Web Scraping Proxy solution with AI Parser.
AI Parser offers 100 free requests per day, making it an ideal starting point for small projects and individual developers. With the free tier, you can experience the power of AI-assisted data extraction without any payment.
AI Parser brings the power of large language models (LLMs) to the web data extraction domain. Unlike traditional scraping tools, it does not depend on the page's DOM structure or CSS selectors. The system analyzes the HTML content you send along with your natural language command.
The LLM understands the page context, determines which data is being requested, and extracts it in a structured format. When you specify a JSON Schema, AI Parser guarantees that the output conforms to that schema — ensuring consistent and predictable data structures.
Extracting product information from e-commerce sites: Pull product name, price, description, stock status, variants, and images with a single command. Collecting article information from news sites: Automatically convert headline, author, date, content, and category into structured format.
Extracting detailed data from listing sites: Systematically extract all features from real estate, vehicle, or job listings. Gathering information from social media profiles: Get username, bio, follower count, and posts in JSON format.
Write your prompts as specifically as possible. Detailed commands like "Extract product name, SKU, price, and stock quantity for each row in the product table" yield better results than "Get the data from the page." Use JSON Schema to standardize output format.
The AI Parser API is designed with a RESTful architecture. It works with simple HTTP POST requests and can be easily called from all popular programming languages. For large HTML pages, sending only the relevant section shortens processing time and increases accuracy.
Experiment with different LLM models to determine which gives the best results for your specific use case. Your extraction commands continue working even when the site structure changes — AI Parser adapts by understanding the page semantics.
Yes, AI Parser offers 100 free requests per day. No credit card required. The free tier is completely sufficient for small projects and testing purposes. Paid plans are available for higher volumes.
AI Parser currently supports GPT-4o and Claude models. Each model has different strengths for different scenarios. You can specify your preferred model when sending a request.
Simply describe what you want to extract in everyday language. For example: "Get the product name, price, and stock status from this page" or "List all reviews with author name and date." You can write in English or any supported language.
No, JSON Schema is optional. If you don't specify a Schema, AI Parser returns the output in the most appropriate structure. However, using Schema is recommended if you want consistent and predictable output.
AI Parser can extract data from e-commerce product pages, news articles, listing sites, social media profiles, forums, catalogs, and many more page types. Any page with HTML content is supported.
You can send 100 requests per day on the free tier. Paid plans offer scalable limits based on your needs. When the rate limit is exceeded, an HTTP 429 status code is returned.
Both methods are supported. When you send a URL, AI Parser fetches the page for you. Sending HTML directly is faster and allows you to process pages fetched with your own proxy infrastructure.
Average response time is between 2-8 seconds. This may vary depending on the page size, prompt complexity, and selected LLM model.
Since AI Parser is offered as a RESTful API, it can be used with any programming language that can send HTTP requests. Python, JavaScript, PHP, Go, Ruby, Java, and all other languages are supported.
AI Parser ensures high accuracy by leveraging the contextual understanding capabilities of LLM models. You can enforce output format with JSON Schema, and experiment with different models to find the one that gives the best results.
Discover our powerful API solution for programmatically fetching HTML pages.
Extract structured data from search engine result pages.
Proxy infrastructure optimized for your web scraping projects.
High success rate data extraction with real user IP addresses.
Proxy solution optimized for SEO tools and analytics platforms.