Hugging Face Model Search & Stats Scraper (Free)

Search the Hugging Face Hub and export clean, structured stats for every model: downloads, likes, trending score, task (pipeline tag), library, license, base models, training datasets, arXiv papers, file count and last update date.

Use it to track AI model popularity, monitor what a competitor or organization is shipping, benchmark model adoption over time, or build a dataset of the LLM/AI ecosystem. Search by keyword or fetch exact model IDs. This actor is 100% free — you only pay your own Apify platform usage.

Input

Field	Type	Default	Description
queries	array of strings	—	Keywords to search the Hub. One item per model found.
modelIds	array of strings	—	Exact repo IDs, e.g. google-bert/bert-base-uncased.
maxResultsPerQuery	integer	10	Models returned per search query (1–100).
sort	string	downloads	downloads, likes, lastModified, createdAt or trendingScore.
direction	string	desc	Sort direction.
pipelineTag	string	""	Optional task filter, e.g. text-generation.
library	string	""	Optional library filter, e.g. transformers.
author	string	""	Optional author filter, e.g. mistralai.

At least one of queries or modelIds is required.

Output

One dataset item per model, always with an explicit status field (found, not_found or error) — the actor never crashes on an unknown model.

{
  "status": "found",
  "query": "bert",
  "source": "search",
  "rank": 1,
  "modelId": "google-bert/bert-base-uncased",
  "author": "google-bert",
  "modelName": "bert-base-uncased",
  "downloads": 78962390,
  "likes": 2704,
  "pipelineTag": "fill-mask",
  "libraryName": "transformers",
  "license": "apache-2.0",
  "languages": ["en"],
  "datasets": ["bookcorpus", "wikipedia"],
  "baseModels": [],
  "arxivIds": ["1810.04805"],
  "tags": ["transformers", "pytorch", "fill-mask"],
  "gated": false,
  "private": false,
  "fileCount": 16,
  "createdAt": "2022-03-02T23:29:04.000Z",
  "lastModified": "2024-02-19T11:06:12.000Z",
  "url": "https://huggingface.co/google-bert/bert-base-uncased",
  "error": null
}

Use cases

• Track AI model trends — snapshot downloads and likes on a schedule to build a time series.
• Competitive & ecosystem research — see every model an organization has published, with licences.
• Model selection — compare candidates for a task by adoption, licence and last update before committing.
• Licence & compliance audits — flag gated or non-commercial models before they reach production.

Limitations & fair use

• Uses the public Hugging Face Hub API. Only public models are returned; private or gated-restricted repos require auth and are out of scope.
• downloads reflects the Hub's rolling 30-day count, not all-time, unless downloadsAllTime is provided by the API.
• Search relevance and ordering are decided by the Hub itself.
• A polite delay is applied between requests. Network calls retry up to 3 times with backoff; failures are reported as an error item rather than crashing the run.
• Model content and licences belong to their respective authors — always check a model's licence before use.

More actors by this developer

Browse the full collection of free data and enrichment actors: apify.com/fit_melon