🤖 AI Lead Extractor

Extract structured lead intelligence — emails, phone numbers, people, company details, and any custom fields you describe — from any website. Every page is run through a fast deterministic contact extractor, and, when you enable AI, an additional structured‑data pass that follows your plain‑English instructions.

---

✨ What it does

• Contact extraction on every page — emails, phone numbers, and social profiles (LinkedIn, X/Twitter, Facebook, Instagram, YouTube, TikTok, GitHub, Discord, Telegram, WhatsApp), plus links, images, and headings. This always runs, with or without AI.
• AI extraction (optional) — describe what you want in plain English and get back structured JSON: company profile, people and roles, or any custom shape you ask for.
• Screenshots (optional) — a JPEG of each page, stored and linked from the result.
• Predictable per‑result pricing — you pay per page that returns data, not per minute of runtime. Pages that fail to load are never charged.
• Proxy support — route requests through Apify Proxy or your own proxy URLs.

---

🚀 Quick start

1. Basic contact extraction (no AI)

{
  "startUrls": [{ "url": "https://example.com/contact" }],
  "useAI": false
}

Returns emails, phones, social links, and page structure.

2. AI extraction with the default lead prompt

{
  "startUrls": [{ "url": "https://apify.com/about" }],
  "useAI": true
}

Returns a structured company/people/contacts object in addition to the basic contacts.

3. AI extraction with your own instructions

{
  "startUrls": [{ "url": "https://apify.com/about" }],
  "useAI": true,
  "aiInstructions": "Extract the company name, the CEO, and the main support email"
}

The AI shapes the output around your request.

---

📥 Input

Field	Type	Required	Default	Description
startUrls	Array	Yes	—	Pages to process. Each entry is { "url": "https://…" }.
useAI	Boolean	No	true	Run the AI extraction pass. When false, only the contact extractor runs.
aiInstructions	String	No	Full lead extraction	Plain‑English description of what to extract. Max 500 characters / 75 words.
includeScreenshot	Boolean	No	true	Capture a JPEG of each page. Turn off if you only need data.
proxyConfig	Object	No	No proxy	Proxy settings (see Proxy).

Default AI instructions (used when useAI: true and aiInstructions is left blank):

Extract all the useful information — email, phone number, description, and any relevant structured data from the page.

---

📤 Output

One dataset item per URL. basicExtraction is always present; aiExtraction appears when AI runs successfully.

{
  "url": "https://apify.com/about",
  "title": "About · Apify",
  "scrapedAt": "2026-07-20T14:30:00.000Z",
  "ok": true,
  "error": null,
  "extractedBy": ["regex", "ai"],
  "basicExtraction": {
    "emails": ["hello@apify.com"],
    "phones": ["+420123456789"],
    "socialLinks": {
      "linkedin": "https://linkedin.com/company/apify",
      "twitter": "https://twitter.com/apify",
      "facebook": null,
      "instagram": null,
      "youtube": null,
      "tiktok": null,
      "github": "https://github.com/apify",
      "discord": null,
      "telegram": null,
      "whatsapp": null
    },
    "links": [{ "text": "Careers", "url": "https://apify.com/jobs", "external": false }],
    "images": [{ "alt": "logo", "url": "https://…/logo.png" }],
    "headings": [{ "level": 1, "text": "About Apify" }],
    "wordCount": 1240,
    "linkCount": 37
  },
  "aiExtraction": {
    "company": { "name": "Apify", "description": "…", "industry": "Software", "address": null, "website": "https://apify.com" },
    "people": [{ "name": "…", "title": "CEO", "email": null, "linkedin": null }],
    "emails": ["hello@apify.com"],
    "phones": ["+420123456789"],
    "socialLinks": { "linkedin": "…", "twitter": "…", "other": [] }
  },
  "content": {
    "markdown": "# About Apify …",
    "screenshot": {
      "available": true,
      "key": "screenshot-1779998222295-ab12cd",
      "url": "https://api.apify.com/v2/key-value-stores/…/records/screenshot-1779998222295-ab12cd",
      "sizeKb": 18
    }
  },
  "metadata": {
    "aiUsage": { "attempted": true, "inputTokens": 2600, "outputTokens": 220, "totalTokens": 2820 }
  }
}

Field reference

Field	Description	Always present
url, title, scrapedAt	Source URL, page title, timestamp	✅
ok	true if the page loaded	✅
error	null, or { code, message, id } when the page failed	✅
extractedBy	["regex"], or ["regex","ai"] when AI succeeded	✅
basicExtraction	Contacts, social links, links, images, headings, counts	✅
aiExtraction	AI structured data (shape follows your instructions); null when AI is off, or an { error } object if the AI pass failed	Only with AI
content.markdown	Cleaned page text	✅
content.screenshot	{ available, key, url, sizeKb }, or { available: false }	✅
metadata.aiUsage	Token accounting for the AI pass	✅

When AI is enabled but the AI pass fails, the page still returns full basicExtraction, aiExtraction carries a sanitized error, and the page is billed at the basic rate — never the AI rate.

---

💰 Pricing & billing

This Actor uses pay‑per‑event pricing: you are charged per page that returns data, at a flat rate per result. Runtime and memory are not billed separately, so cost is predictable regardless of how long a page takes.

Two events:

Event	When	FREE	BRONZE	SILVER	GOLD
AI Extraction	useAI: true and the AI pass succeeds	$0.045	$0.018	$0.015	$0.012
Basic Extraction	useAI: false, or AI enabled but its pass failed	$0.030	$0.010	$0.009	$0.008

Prices are per page. Paid Apify plans (Bronze → Gold) receive progressively lower rates.

What you are not charged for:

• Pages that fail to load — no result, no charge.
• The AI pass failing — that page drops to the Basic rate.

AI model usage is separate and billed to your own Apify account, at cost, when AI is enabled. On the free Apify plan, AI responses are shorter (capped around 2,048 tokens) and cost more per token; paid Apify plans get full‑length responses at lower cost. This model usage is typically a fraction of a cent per page.

Spending cap: set a maximum charge per run (ACTOR_MAX_TOTAL_CHARGE_USD) and the Actor stops once it's reached, finishing the page in progress first.

---

🌍 Proxy

Requests can be routed through a proxy to reduce rate‑limiting and reach region‑specific content. Configure it under proxyConfig:

{
  "startUrls": [{ "url": "https://example.com" }],
  "proxyConfig": { "useApifyProxy": true }
}

• Apify Proxy — enable useApifyProxy. A fresh IP is drawn per page.
• Your own proxies — supply proxyUrls with your proxy endpoints; requests route through them as‑is.
• Residential exit IPs — if you need residential IPs, provide them through your own proxyUrls.

If a proxy can't be applied, the run continues on a direct connection rather than failing.

---

🧠 Memory

The Actor runs a real browser, so it needs 1 GB (the default). You can raise it to 2 GB for heavier pages; because billing is per result, extra memory changes speed, not the per‑page price.

---

🎯 Use cases

• Lead generation — pull contacts and decision‑makers from company, team, and about pages.
• Contact enrichment — turn a list of company URLs into emails, phones, and social profiles.
• Competitive research — capture structured company profiles and screenshots at scale.
• Custom extraction — describe any structured output ("list job openings with title and location", "extract pricing tiers and features") and get it as JSON.

---

❓ FAQ

Do I have to use AI? No. With useAI: false you get the full contact extractor (emails, phones, socials, links, headings) at the Basic rate.

What happens if a page blocks me? That page returns ok: false with a sanitized error and is not charged. The run continues to the next URL.

Can I control the output shape? Yes — aiInstructions drives the structure of aiExtraction. Leave it blank for the full lead‑extraction schema.

Why did a page get the Basic rate when I enabled AI? The AI pass didn't complete for that page (for example, a model limit on the free plan). You still get full contact data, so the page is billed at the lower Basic rate.

How do I keep costs bounded? Set a max charge per run, disable screenshots if you don't need them, and start at 1 GB memory.

---

🆘 Support

Open an issue on the Actor's Apify page, or contact the developer through Apify Store.