It's an AI tool that automatically figures out what a business does (its industry) just by looking at its website. It saves sales and marketing teams hours of manual research.

You give it a list of websites (in a CSV file). It visits each site, reads the text, and uses advanced AI (like GPT-4, Claude, Gemini) to determine the most likely industry classifications (e.g., SIC, NAICS codes) and provides a confidence score.

Just a CSV file with a column containing website domains (e.g., 'example.com'). More columns like company name or existing description can help improve accuracy.

You get your original CSV file back, but with new columns added: Industry, Sub-Industry, SIC Code, NAICS Code, Confidence Score, and potentially enriched data like company description, keywords, etc.

Version 8.5 supports OpenAI (GPT-4o mini, GPT-4 Turbo), Anthropic (Claude 3.5 Sonnet, Opus), and Google Gemini Pro. It can automatically choose the best one or fallback if one fails.

It's a point-and-click tool included in HuntLeads.ai. You load a website, click on the data you want (like prices, names, addresses), and it learns how to grab that data automatically. You can save these instructions as templates – no coding needed!

Yes, the data enrichment feature attempts to find contact information like emails and phone numbers listed on the website, as well as social media profiles and key personnel if available.

HuntLeads.ai offers a user-friendly Desktop App (Windows, Mac, Linux), a powerful Command-Line Interface (CLI) for automation, and a REST API for integration into other systems.

Yes! Version 8.5 includes a real-time web dashboard (usually runs on port 5000) where you can monitor job progress, view system stats (CPU/Memory), and see logs as they happen.

Yes, you need to provide your own API key(s) for the AI provider(s) you want to use (OpenAI, Anthropic, Google). HuntLeads.ai securely uses your keys to make classification requests.

Yes, HuntLeads.ai is designed to run locally on your machine (Windows, Mac, Linux) or on your own servers. An internet connection is needed for the AI API calls.

Thanks to the auto-resume feature, you can simply restart the process (using the `--resume` flag in CLI or via the GUI). It will pick up where it left off using saved checkpoints, so you don't lose progress or waste API calls.

Our 4-stage classification algorithm achieves 95%+ accuracy. We use three different AI providers (Claude, GPT-4, Gemini) and cross-reference against our 50,000+ service category database plus SIC/NAICS codes. Every lead gets a confidence score so you can filter by quality.

Our triple-redundant architecture automatically fails over to the next provider in milliseconds. Claude → GPT-4 → Gemini. You'll never experience downtime. This is a huge advantage over competitors who rely on a single AI.

Absolutely! We offer a full REST API, CLI tool, webhooks, and direct integrations with popular platforms like Salesforce, HubSpot, and more. You can also export to any database (PostgreSQL, MySQL, MongoDB) or cloud storage (S3, Azure, GCS).

With 8-12 parallel workers, we process up to 3,600 leads per minute. That's 216,000 leads per hour or 5.2 million leads per day at maximum capacity. Most customers process their entire database in hours, not weeks.

Preview lets you test your selectors before running a full scrape. It extracts up to 5 rows entirely in your browser — no server calls, no webhooks, no database writes. Results appear with a blue banner so you can distinguish them from production scrapes. Your last full scrape export (CSV/JSON) stays intact and accessible.

Page Watcher monitors public URLs on schedules you set (daily, twice-daily, weekly). It detects content changes and pushes alerts via WebSocket to the Chrome extension. GoLiveScore is a 0-100 heuristic that scores each domain across 5 factors (content depth, structural signals, parking indicators, business signals, tech maturity) to answer: "Has this parked domain become a real business?" Scores above 75 are classified as "Live" — your signal to reach out.

The scoring engine evaluates 8 weighted factors: email validity, phone validity, employee count, estimated revenue, website existence, social presence, industry fit, and engagement signals. Each factor contributes to a 0-100 composite score. Leads are ranked into four tiers — Hot (80-100), Warm (60-79), Cold (40-59), and Nurture (0-39). You can adjust factor weights via sliders to match your ideal customer profile.

HuntLeads includes a full data quality suite: Smart Deduplication (key-based or full-row, first-wins, case-insensitive), Output Normalization (auto-formats phones, URLs, emails, whitespace), Cell Validation (flags invalid emails, malformed phones, and broken URLs with red indicators), an Email Verifier (MX/SMTP/catch-all checks), and a Duplicates Manager (multi-criteria matching across email, phone, domain, and fuzzy company name). All of these run automatically in the scraper pipeline and are also available standalone.

Flat/Single Page: Extracts all matching elements from the current viewport at once — fastest, no server calls, and Preview-supported. Paginated: Follows "Next" buttons across multiple pages via SSE streaming with configurable delays and page limits. Click-Each: Opens each detail page from a listing, extracts fields, navigates back, and streams results — perfect for directories. Follow-Links: Extracts listing data AND follows each link to merge detail-page fields into enriched rows.

The Free B2B Leads Chrome Extension lives in Chrome's persistent Side Panel. Browse any website and the extension auto-detects the domain, enriches it with 8 data suppliers, and shows company data, contacts, tech stack, and a lead score. One click on "Push to Hunt Leads" (or Ctrl+Shift+H) sends the domain for deep scraping, business classification, and GoLiveScore analysis. A "Known Lead" badge prevents double-prospecting. The extension is included free with any HuntLeads.ai or M365 Cockpit license.

Yes. The Proxy Manager lets you configure and rotate residential HTTP/HTTPS and SOCKS5 proxies. Built-in health checks verify connectivity before each request, and automatic rotation distributes load across your pool. Performance analytics show response times, success rates, and usage per proxy. You can test connectivity with one click from the settings page.

Yes. Every feature is accessible via a RESTful JSON API authenticated with API keys. Endpoints cover scrape execution (all four modes), business classification, lead scoring, page watch management (CRUD + status), batch upload, webhook registration, and data export (CSV, JSON). The API supports SSE streaming for real-time scrape progress and is built on FastAPI with async I/O for maximum throughput.

Yes. An APScheduler-powered scheduler runs watches and batch tasks on fixed UTC cron schedules — daily, twice-daily, or weekly. Each watch can be paused and resumed independently without deletion. Scheduled checks trigger GoLiveScore updates and optional webhook notifications when content changes or score thresholds are crossed.

CRM Relay handles Salesforce OAuth server-side so the Chrome Extension never touches credentials. When you push leads from the extension or from the scraper results table, Hunt Leads upserts records into Salesforce by email address — preventing duplicates. Field mapping is configurable from the settings page. HubSpot and Pipedrive integrations use the same relay pattern with their respective API conventions.

HuntLeads.ai supports CSV, Excel (XLSX/XLS), JSON, and TSV file formats for both input and output. You can also integrate directly with databases (PostgreSQL, MySQL, MongoDB) and cloud storage (AWS S3, Azure Blob, Google Cloud Storage).

Yes! HuntLeads.ai allows you to define custom classification schemas with your own categories and subcategories. You can also configure scoring criteria and validation rules to match your specific business needs.

The Visual Scraper is a point-and-click tool that lets you select elements on any webpage to extract data. It learns the directory structure and can scrape entire sites like Google Maps, Yellow Pages, LinkedIn, Yelp, and more. No coding required!

The GUI (Desktop App) provides a visual interface with drag-and-drop, real-time progress tracking, and interactive results. The CLI (Command Line) is perfect for automation and batch scripting. Both offer the same core functionality and run entirely on your machine.

Absolutely! You can use the CLI with cron jobs (Linux/Mac) or Task Scheduler (Windows) to run classifications on a schedule. The REST API also supports webhook integration for event-driven automation.

Our smart routing system automatically distributes requests across OpenAI, Claude, and Gemini based on availability, performance, and cost. If one provider is slow or unavailable, requests are instantly routed to the next provider with zero downtime.

HuntLeads.ai classifies businesses across ALL industries including B2B Services (Technology, Professional Services, Consulting), B2C (Retail, Healthcare, Food & Beverage), Manufacturing, Real Estate, Finance, Education, and more. We support 50,000+ service categories plus SIC/NAICS codes.

Yes! HuntLeads.ai integrates directly with Salesforce, HubSpot, Pipedrive, and other major CRMs via our REST API and webhooks. You can also export to CSV/Excel and import manually.

No limit! We've tested with batches of 10+ million leads. The system uses smart checkpointing and auto-resume, so even if processing is interrupted, you can pick up exactly where you left off without wasting API calls.

You pay directly to the AI providers (OpenAI, Anthropic, Google) based on your usage. Typical costs range from $0.50 to $20 per 1,000 leads depending on which AI model you choose. GPT-3.5 and Claude Haiku are most cost-effective, while GPT-4 and Claude Opus offer highest accuracy.

The core application runs locally on your machine, but you'll need an internet connection to make AI API calls for classification. Once classified, you can work with the results completely offline.

Each classification includes a confidence score (0-100%). Scores above 90% are highly reliable, 70-89% are good quality, and below 70% may need manual review. You can filter and sort results by confidence to prioritize your best leads.

Yes! All data is processed locally on your machine or servers. We don't store or transmit your lead data to our servers. You maintain complete control over your data, ensuring GDPR, CCPA, and other privacy compliance.

HuntLeads.ai automatically handles rate limiting with smart retry logic and exponential backoff. If one provider hits rate limits, the system immediately switches to another provider so your processing continues without interruption.