Hugging Face Trending Models, Datasets & AI Papers Radar

Extract real-time trending AI models, high-velocity datasets, daily research papers, download metrics, and licensing metadata directly from Hugging Face without requiring API tokens or authentication keys.

Built natively for AI Agents (Model Context Protocol / MCP), machine learning researchers, tech venture scouts, developers, and AI content curators.

---

🚀 Key Capabilities

• 🔥 Trending AI Models: Monitor breakout open-source models across Transformers, LLMs, Vision, Audio, Multimodal, and Diffusion.
• 📦 Datasets Discovery: Uncover trending pre-training corpora, instruction-tuning datasets, and evaluation benchmarks.
• 📄 Daily AI Research Papers: Track breakthrough papers submitted to ArXiv with community upvotes and abstracts.
• 📊 Real-time Velocity & Engagement: Capture live trendingScore, total likes, monthly downloads, and commit timestamps.
• 🏷️ Deep Tag & License Extraction: Identify model architectures (safetensors, endpoints_compatible, 8-bit, fp8) and explicit software licenses (mit, apache-2.0, llama3).
• ⚡ Zero Authentication Required: Uses public high-throughput endpoints with zero rate-limit hurdles or login barriers.
• 🤖 MCP & AI Agent Native: Structured JSON schema ready for instant ingestion into Claude Desktop, Cursor, ChatGPT, and LangChain pipelines.

---

💼 Ideal Use Cases

1. AI Market Intelligence & VC Scouting: Track breakout open-weight releases before they go viral on X or Hacker News.
2. AI Agent Tooling (MCP): Connect your autonomous AI coding agents to dynamically query state-of-the-art models for specific tasks (e.g. image-text-to-text, text-generation).
3. Daily AI Newsletters & Curators: Automatically compile daily or weekly roundups of top models and research papers.
4. License & Compliance Auditing: Monitor model licensing types (mit, apache-2.0, commercial-friendly) for enterprise adoption.

---

📥 Input Parameters

Parameter	Type	Default	Description
resourceType	String	"models"	What to extract: models, datasets, or papers.
sortBy	String	"trendingScore"	Sort criterion: trendingScore, likes, downloads, createdAt, lastModified.
searchQuery	String	""	Optional keyword filter (e.g. deepseek, llama, flux, whisper).
authorFilter	String	""	Filter by organization/author (e.g. deepseek-ai, meta-llama, google).
taskFilter	String	""	Filter models by pipeline task (e.g. text-generation, image-text-to-text).
maxResults	Integer	30	Number of items to extract (1 to 200).

---

📤 Output Dataset Schema

Each item pushed to your Apify dataset contains comprehensive metadata:

Field	Type	Description
id	String	Unique repo slug (e.g. deepseek-ai/DeepSeek-V4.1-Flash)
resourceType	String	Type: model, dataset, or paper
title	String	Title or repository ID
author	String	Organization or creator username
likes	Integer	Community upvotes count
trendingScore	Integer	Hugging Face trending velocity score
downloads	Integer	Monthly download count
pipelineTag	String	Machine learning task tag
license	String	Model/dataset software license
tags	Array	Full list of Hugging Face tags
tagsString	String	Comma-separated tag string for spreadsheets
createdAt	String	ISO 8601 creation date
lastModified	String	ISO 8601 update date
url	String	Canonical link to Hugging Face page
summary	String	Descriptive summary or abstract snippet

---

📋 Sample JSON Output

[
  {
    "id": "deepseek-ai/DeepSeek-V4.1-Flash-1017",
    "resourceType": "model",
    "title": "deepseek-ai/DeepSeek-V4.1-Flash-1017",
    "author": "deepseek-ai",
    "likes": 3581,
    "trendingScore": 889,
    "downloads": 116770,
    "pipelineTag": "image-text-to-text",
    "libraryName": "transformers",
    "license": "mit",
    "tags": [
      "transformers",
      "safetensors",
      "deepseek_v41",
      "text-generation",
      "image-text-to-text",
      "license:mit"
    ],
    "tagsString": "transformers, safetensors, deepseek_v41, text-generation, image-text-to-text, license:mit",
    "createdAt": "2026-10-17T02:40:38.000Z",
    "lastModified": "2026-10-28T18:12:00.000Z",
    "private": false,
    "url": "https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017",
    "summary": "Hugging Face AI Model deepseek-ai/DeepSeek-V4.1-Flash-1017 (image-text-to-text) with 3581 likes and 116770 downloads."
  }
]

---

💰 Pricing (Pay-Per-Event)

• Actor Start: $0.00008 per run
• Per Result Item: $0.0034 per extracted item ($3.40 / 1,000 items)
• Zero Hidden Infrastructure Charges: Platform usage compute fees are disabled for clean and transparent pricing.