Trending AI Agents & LLMs

Fetches the latest trending AI/LLM models, datasets, and Spaces from Hugging Face via the official keyless API (https://huggingface.co/api/trending). No scraping, no API key, no auth.

Why this is not a duplicate of github-trending-scraper: that actor scrapes the github.com/trending HTML page (GitHub repositories, by language/date). This actor reads Hugging Face's official trending endpoint — a completely different platform, data model (models/datasets/Spaces vs repos), and delivery mechanism (JSON API vs HTML scrape).

Input

{
  "repoTypes": ["model", "dataset", "space"],
  "limit": 10,
  "includeAuthorData": true,
  "includeInferenceProviders": true
}

• repoTypes — array of model, dataset, space (default all three).
• limit — max trending items per repo type (API exposes up to 10 per type; default 10).
• includeAuthorData — include author profile (name, fullname, account type, follower count).
• includeInferenceProviders — for models, include third-party inference providers hosting the model.

Output

One dataset row per trending item. Common fields: repoType, id, url, author, likes, downloads, lastModified, gated, private. Type-specific extras:

• model — pipelineTag, numParameters, availableInferenceProviders[] (provider, providerId, task, modelStatus)
• dataset — numRows, modalities[], isBenchmark, isTraces
• space — title, emoji, runtime, tags[], aiCategory, aiShortDescription

Failed type fetches produce explicit error rows ({"repoType": ..., "id": null, "error": "..."}).

Pricing

PAY_PER_EVENT — $0.001 per item-scraped event (one per dataset row), apifyMarginPercentage: 0.2.

Local run

cd trending-ai-agents
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
echo '{"repoTypes": ["model", "dataset", "space"], "limit": 3}' > INPUT.json
APIFY_PYTHON_COMMAND=.venv/bin/python apify run --input-file INPUT.json

Publish

Requires an authenticated Apify CLI (apify login with APIFY_TOKEN):

$apify push --force

Then set PAY_PER_EVENT pricing via the Apify API (actor ID comes from the push output):

curl -X PUT "https://api.apify.com/v2/acts/{ACTOR_ID}?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "pricingModel": "PAY_PER_EVENT",
    "pricingPerEvent": {
      "actorChargeEvents": {
        "item-scraped": {"eventPriceUsd": 0.001}
      }
    },
    "apifyMarginPercentage": 0.2,
    "isPublic": true,
    "title": "Trending AI Agents & LLMs",
    "categories": ["AI", "DEVELOPER_TOOLS"]
  }'

Verify monetization with a platform run: POST /v2/acts/{ACTOR_ID}/runs with the raw input object as body, then confirm chargedEventCounts.item-scraped on the completed run.

Related actors

• GitHub Trending Scraper — trending GitHub repos by language and time range
• Hacker News Story Stats — HN front-page and search story stats