🤗 Hugging Face AI Models & Datasets Scraper

Scrape trending machine learning models, download metrics, likes, tasks, licenses, authors, and metadata directly from the Hugging Face Hub without API keys or rate limits.

---

🥊 Why This Actor Beats the Competition

Feature	🤗 Hugging Face Models Scraper	Manual Hub Browsing	Generic Web Scrapers
Speed	200+ models in < 2 seconds	Manual copy-pasting	10–30s with heavy DOM rendering
Pricing	Pay-Per-Result ($1.20 / 1k models)	Free manual labor	Costly headless browser compute
Deep Filtering	By pipeline tag, downloads, likes, trending score	Basic UI search	Fragile CSS selectors
Complete Metadata	Licenses, tags, gating status, commit timestamps	Partial card data	Missing library names
AI Stack Ready	Structured JSON for GPU provisioning & model monitors	Web page only	Unclean HTML tags

---

🛠️ Step-by-Step: How to Use

Method 1: Apify Console (No-Code GUI)

1. Select your target Task / Pipeline Tag (e.g. text-generation, image-to-image, or leave blank for all).
2. Choose a sort order: Most Downloads, Most Likes, or Trending Score.
3. Set your maximum model limit (default 25) and click Start.

Method 2: Python SDK

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run_input = {
    "pipelineTag": "text-generation",
    "sort": "downloads",
    "limit": 20
}

run = client.actor("litezero/huggingface-models-scraper").call(run_input=run_input)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(f"{item['modelId']} - {item['downloads']:,} downloads (License: {item['license']})")

Method 3: Node.js SDK

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_APIFY_TOKEN' });
const run = await client.actor('litezero/huggingface-models-scraper').call({
    pipelineTag: 'text-generation',
    sort: 'trendingScore',
    limit: 10,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Method 4: cURL / REST API

curl --request POST \
  --url "https://api.apify.com/v2/acts/litezero~huggingface-models-scraper/runs?token=YOUR_APIFY_TOKEN" \
  --header "Content-Type: application/json" \
  --data '{"pipelineTag": "text-generation", "limit": 15}'

---

📥 Input Parameters

Parameter	Type	Default	Description
pipelineTag	String	"text-generation"	Machine learning task (e.g. text-generation, automatic-speech-recognition, or all).
sort	Select	"downloads"	Sorting method (downloads, likes, trendingScore, createdAt, lastModified).
searchQuery	String	""	Optional keyword filter (e.g. llama, whisper, deepseek).
limit	Integer	25	Total number of models to retrieve.

---

📤 Sample Output JSON

{
  "modelId": "meta-llama/Meta-Llama-3.1-8B-Instruct",
  "author": "meta-llama",
  "pipelineTag": "text-generation",
  "downloads": 1420580,
  "downloadsAllTime": 4820194,
  "likes": 8412,
  "trendingScore": 96.4,
  "libraryName": "transformers",
  "license": "llama3.1",
  "tags": ["text-generation", "pytorch", "llama", "conversational"],
  "private": false,
  "gated": true,
  "lastModified": "2026-08-15T18:12:00.000Z",
  "modelUrl": "https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct",
  "scrapedAt": "2026-10-02T16:45:00.000Z"
}

---

💡 Top Commercial Use Cases

1. GPU Cloud Intelligence: Track breakthrough models gaining download momentum to optimize hardware deployments.
2. AI Competitive Benchmarking: Monitor enterprise release velocity (Meta, Mistral, Microsoft, Google, DeepSeek).
3. Automated LLM Directory Portals: Power live catalogs of open-source models with up-to-date licenses and metrics.