March 4, 2026 tutorial

WebPerception API: Quickstart Guide — From Zero to Web Data in 5 Minutes

Target keyword: "web scraping API quickstart" (1,200/mo) + "web scraping API tutorial" (1,900/mo)

Word count: ~1,600

---

You have an AI agent that needs to see the web. Maybe it monitors competitor prices, extracts product data, or takes screenshots for visual QA. The problem? Building reliable web scraping infrastructure is a nightmare.

WebPerception API gives your agent eyes on the web — scraping, extraction, and screenshots through three simple endpoints. No browser farms. No proxy rotation. No headless Chrome configs.

Here's how to go from zero to structured web data in under 5 minutes.

Step 1: Get Your API Key (Free)

Once you have your key, set it as an environment variable:

export MANTIS_API_KEY="your-api-key-here"

Step 2: Scrape Your First Page

The /scrape endpoint returns clean, structured content from any URL.

import requests

response = requests.post(
    "https://api.mantisapi.com/scrape",
    headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
    json={"url": "https://news.ycombinator.com"}
)

data = response.json()
print(data["content"])  # Clean text/markdown content
print(data["metadata"])  # Title, description, links

That's it. One call. No Puppeteer setup, no proxy configuration, no CAPTCHA handling. WebPerception handles all of that behind the scenes.

What You Get Back

{
  "content": "Clean markdown or text content of the page...",
  "metadata": {
    "title": "Hacker News",
    "description": "...",
    "url": "https://news.ycombinator.com",
    "links": ["https://...", "https://..."],
    "statusCode": 200
  }
}

Step 3: Extract Structured Data with AI

The /extract endpoint is where it gets powerful. Instead of writing CSS selectors or regex patterns, you describe what you want in plain English.

response = requests.post(
    "https://api.mantisapi.com/extract",
    headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
    json={
        "url": "https://example.com/products/widget-pro",
        "prompt": "Extract the product name, price, rating, and number of reviews"
    }
)

data = response.json()
print(data["extracted"])

Output

{
  "extracted": {
    "product_name": "Widget Pro",
    "price": "$49.99",
    "rating": 4.7,
    "num_reviews": 1243
  }
}

No selectors. No XPath. No breaking when the site redesigns. The AI extraction adapts to any page layout.

Step 4: Take Screenshots

The /screenshot endpoint captures full-page or viewport screenshots — perfect for visual monitoring, QA, or archiving.

response = requests.post(
    "https://api.mantisapi.com/screenshot",
    headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
    json={
        "url": "https://mantisapi.com",
        "fullPage": True
    }
)

# Save the screenshot
with open("screenshot.png", "wb") as f:
    f.write(response.content)

Use cases:

Visual regression testing — compare screenshots before/after deploys
Competitor monitoring — track pricing page changes visually
Compliance archiving — snapshot pages for legal/regulatory records

Step 5: Build an Agent with WebPerception

Here's a real-world example: a competitive intelligence agent that monitors competitor pricing daily.

import requests
import json

MANTIS_API_KEY = "your-api-key"
COMPETITORS = [
    "https://competitor-a.com/pricing",
    "https://competitor-b.com/pricing",
    "https://competitor-c.com/pricing"
]

def check_competitor_prices():
    results = []
    for url in COMPETITORS:
        response = requests.post(
            "https://api.mantisapi.com/extract",
            headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
            json={
                "url": url,
                "prompt": "Extract all pricing tiers with plan name, monthly price, and key features"
            }
        )
        results.append({
            "url": url,
            "pricing": response.json()["extracted"]
        })
    return results

# Run daily via cron, send results to Slack/email
prices = check_competitor_prices()
print(json.dumps(prices, indent=2))

Three competitors. Three API calls. Structured pricing data. No scraping infrastructure to maintain.

Pricing

|------|------------|-------|---------|

| Free | 100 | $0 | $0 |

| Starter | 5,000 | $29/mo | $0.0058 |

| Pro | 25,000 | $99/mo | $0.0040 |

| Scale | 100,000 | $299/mo | $0.0030 |

Overage: $0.005 per additional call on any paid plan.

Using WebPerception with Popular Frameworks

LangChain

from langchain.tools import tool
import requests

@tool
def web_scrape(url: str) -> str:
    """Scrape a webpage and return its content."""
    response = requests.post(
        "https://api.mantisapi.com/scrape",
        headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
        json={"url": url}
    )
    return response.json()["content"]

@tool
def web_extract(url: str, query: str) -> dict:
    """Extract structured data from a webpage using natural language."""
    response = requests.post(
        "https://api.mantisapi.com/extract",
        headers={"Authorization": f"Bearer {MANTIS_API_KEY}"},
        json={"url": url, "prompt": query}
    )
    return response.json()["extracted"]

OpenAI Function Calling

tools = [
    {
        "type": "function",
        "function": {
            "name": "web_scrape",
            "description": "Scrape a webpage and return clean content",
            "parameters": {
                "type": "object",
                "properties": {
                    "url": {"type": "string", "description": "URL to scrape"}
                },
                "required": ["url"]
            }
        }
    },
    {
        "type": "function",
        "function": {
            "name": "web_extract",
            "description": "Extract structured data from a webpage",
            "parameters": {
                "type": "object",
                "properties": {
                    "url": {"type": "string", "description": "URL to extract from"},
                    "prompt": {"type": "string", "description": "What data to extract"}
                },
                "required": ["url", "prompt"]
            }
        }
    }
]

What's Next?

Sign up free — 100 calls/month, no credit card

Read the API docs — full endpoint reference

Join the community — share what you're building, get help from other agent developers

---

WebPerception API is built by Mantis — web perception infrastructure for AI agents. Give your agents eyes on the web.

Ready to try Mantis?

100 free API calls/month. No credit card required.

Get Your API Key →