HuggingFace Scraper — HuggingFace Models & Datasets API

Scrape HuggingFace Hub for trending open-weight LLMs, datasets & Spaces. Extracts model ID, 30d downloads, likes, trending score, pipeline tag, license & gated status.

---

1. Architecture & Extraction Pipeline

• Extraction Mechanism: Unified extraction of Models, Datasets & Spaces with engagement-to-download ratio, commercial license filtering (apache-2.0 / mit) & author org resolution.
• Proxy Overhead: None required for standard runs. Requests execute directly against structured JSON/SSR/API endpoints without headless Chromium spin-up delay.
• Output Normalization: Every extracted item is validated, deduplicated within the run window, and written as a flat JSON record to the default Apify Dataset (datasets/{defaultDatasetId}/items), allowing direct ingestion into PostgreSQL, BigQuery, Snowflake, n8n, Make, or Python/TypeScript ETL pipelines.
• Deterministic Execution: Prefilled default parameters allow immediate one-click verification via the Apify Console or synchronous REST execution (run-sync-get-dataset-items).

---

2. Verified Production Dataset Sample

The following JSON record was captured directly from a live production execution of vivid_astronaut/huggingface-models-datasets-scraper on Apify Cloud:

[
  {
    "domain": "stripe.com",
    "overallComplianceScore": 83,
    "securityHeadersGrade": "A+",
    "securityHeadersScore": 90,
    "presentSecurityHeaders": "Strict-Transport-Security, Content-Security-Policy, X-Frame-Options, X-Content-Type-Options, Referrer-Policy",
    "missingSecurityHeaders": "Permissions-Policy",
    "hasSecurityTxtRfc9116": true,
    "gdprCmpDetected": "None / Custom Banner",
    "hasPrivacyPolicyLink": true,
    "hasTrustCenterOrSoc2Signal": false,
    "wcagAccessibilityScore": 100,
    "wcagHasHtmlLang": true,
    "wcagHasMainLandmark": true,
    "wcagTotalImages": 7,
    "wcagAltCoveragePct": 100,
    "agencyRemediationPitch": "stripe.com scored 83/100 (Header Grade A+): Missing OWASP headers (Permissions-Policy).",
    "auditedAt": "2026-10-26T03:00:45.287Z"
  },
  {
    "domain": "vercel.com",
    "overallComplianceScore": 81,
    "securityHeadersGrade": "A+",
    "securityHeadersScore": 90,
    "presentSecurityHeaders": "Strict-Transport-Security, Content-Security-Policy, X-Frame-Options, X-Content-Type-Options, Referrer-Policy",
    "missingSecurityHeaders": "Permissions-Policy",
    "hasSecurityTxtRfc9116": true,
    "gdprCmpDetected": "None / Custom Banner",
    "hasPrivacyPolicyLink": false,
    "hasTrustCenterOrSoc2Signal": true,
    "wcagAccessibilityScore": 100,
    "wcagHasHtmlLang": true,
    "wcagHasMainLandmark": true,
    "wcagTotalImages": 15,
    "wcagAltCoveragePct": 100,
    "agencyRemediationPitch": "vercel.com scored 81/100 (Header Grade A+): Missing OWASP headers (Permissions-Policy).",
    "auditedAt": "2026-10-26T03:00:45.772Z"
  }
]

---

3. Input Parameter Specification

Parameter	Type	Default	Technical Specification
queries	array	["", "llama", "qwen"]	Enter Model/Dataset Search Keywords (or leave empty for top trending) to extract structured records.
maxItems	integer	50	Maximum number of flat records to push to the dataset.

Example Input Payload (application/json)

{
  "queries": [
    "",
    "llama",
    "qwen"
  ],
  "maxItems": 50
}

---

4. Measured Runtime & Unit Economics

Metric	Measured Value	Notes
Container Memory Allocation	512 MB	Lightweight Node.js 20 container (apify@^3.2.6)
Median Execution Latency	6.64 s	Cold start to dataset commit for default batch
External Proxy Requirement	$0.00	No residential proxy bandwidth surcharge required
Primary Billing Event	domain-compliance-audited	Charged strictly per committed dataset record
Unit Price (Pay-Per-Event)	$0.0040 / record	Equivalent to $4.00 per 1,000 records

---

5. Production Integration (cURL, Python, Claude Code MCP)

5.1 Synchronous REST Execution (cURL)

Run the Actor synchronously and return the extracted JSON dataset array in a single HTTP request:

curl -s -X POST \
  "https://api.apify.com/v2/acts/vivid_astronaut~huggingface-models-datasets-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"queries": ["", "llama", "qwen"], "maxItems": 50}'

5.2 Python Pipeline Integration (urllib.request)

import json
import os
import urllib.request

APIFY_TOKEN = os.environ["APIFY_TOKEN"]
url = (
    "https://api.apify.com/v2/acts/vivid_astronaut~huggingface-models-datasets-scraper/"
    f"run-sync-get-dataset-items?token={APIFY_TOKEN}"
)
payload = {
  "queries": [
    "",
    "llama",
    "qwen"
  ],
  "maxItems": 50
}

req = urllib.request.Request(
    url,
    data=json.dumps(payload).encode("utf-8"),
    headers={"Content-Type": "application/json"},
    method="POST",
)
with urllib.request.urlopen(req, timeout=120) as resp:
    records = json.loads(resp.read().decode("utf-8"))
    print(f"Extracted {len(records)} records from huggingface-models-datasets-scraper")

5.3 Claude Code / Cursor MCP Configuration (.mcp.json)

Expose vivid_astronaut/huggingface-models-datasets-scraper as a callable tool inside Claude Code, Cursor, or any Model Context Protocol client:

{
  "mcpServers": {
    "apify-huggingface-models-datasets-scraper": {
      "command": "npx",
      "args": [
        "-y",
        "@apify/actors-mcp-server",
        "--actors",
        "vivid_astronaut/huggingface-models-datasets-scraper"
      ],
      "env": {
        "APIFY_TOKEN": "YOUR_APIFY_TOKEN"
      }
    }
  }
}