Hugging Face Model Scraper

Extract comprehensive data from the Hugging Face model hub -- the world's largest repository of AI and machine learning models. This scraper uses the official Hugging Face API for fast, reliable data extraction.

What does Hugging Face Model Scraper do?

This actor scrapes model metadata from Hugging Face, including:

• Model ID -- Full identifier (e.g., meta-llama/Llama-2-7b)
• Author -- Organization or user who published the model
• Downloads -- Total download count
• Likes -- Community likes/stars
• Task -- Pipeline task type (text-generation, image-classification, etc.)
• Tags -- All model tags (framework, language, license, etc.)
• Library -- ML framework used (transformers, diffusers, etc.)
• Last Modified -- When the model was last updated
• Model URL -- Direct link to the model page

How to use

Search by query

Provide a searchQuery to find models matching a keyword. For example, searching for "text-generation" returns the most popular text generation models sorted by downloads.

Scrape specific models

Provide direct Hugging Face model URLs in the urls field to extract detailed data for specific models.

Input configuration

Field	Type	Default	Description
searchQuery	String	text-generation	Search term to find models
urls	Array	[]	Direct model page URLs to scrape
maxResults	Integer	100	Maximum number of models to extract (1-1000)
useResidentialProxy	Boolean	false	Use residential proxy for better success rates

Example input

{
    "searchQuery": "stable-diffusion",
    "maxResults": 50
}

Example input with direct URLs

{
    "urls": [
        { "url": "https://huggingface.co/meta-llama/Llama-2-7b" },
        { "url": "https://huggingface.co/openai/whisper-large-v3" }
    ]
}

Output example

{
    "modelId": "meta-llama/Llama-2-7b-chat-hf",
    "author": "meta-llama",
    "modelName": "Llama-2-7b-chat-hf",
    "downloads": 15234567,
    "likes": 4521,
    "task": "text-generation",
    "tags": ["transformers", "pytorch", "llama", "text-generation", "en"],
    "lastModified": "2024-03-15T10:30:00.000Z",
    "libraryName": "transformers",
    "private": false,
    "modelUrl": "https://huggingface.co/meta-llama/Llama-2-7b-chat-hf",
    "scrapedAt": "2026-02-11T12:00:00.000Z"
}

Use cases

• AI Research -- Track trending models and compare download metrics
• ML Engineering -- Find the best models for specific tasks
• Market Analysis -- Monitor AI model ecosystem and framework adoption
• Competitive Intelligence -- Track organizations publishing models
• Dataset Building -- Collect model metadata for analysis

Performance

• Uses the official Hugging Face API for fast, structured JSON responses
• Supports pagination for large result sets
• Typical run: 100 models in under 30 seconds
• Low memory footprint (256-512 MB)

Cost estimation

Using Apify platform pricing:

• ~100 models: ~$0.001
• ~1,000 models: ~$0.01
• Pay-per-event pricing: $0.0003 per result

Integrations

Connect Hugging Face Model Scraper with your favorite tools:

• Export to Google Sheets, Excel, CSV, or JSON
• Send data to webhooks or APIs
• Integrate with Zapier, Make, or other automation platforms
• Store results in Apify datasets for further processing

Technical details

• Built with Apify SDK v3 and Crawlee
• Uses CheerioCrawler for lightweight HTTP requests
• Fetches data from Hugging Face REST API (/api/models)
• No browser needed -- fast and efficient