🔎 Open LLM Leaderboard Scraper

🚀 Scrape the Hugging Face Open LLM Leaderboard.

🕒 Last updated 2026-07-17 · 📊 12 fields per record · no API key needed

Pull model rankings, benchmark scores (ARC, HellaSwag, MMLU, TruthfulQA, Winogrande, GSM8K), parameter counts, precision, and submitter. Every record is mapped to a consistent schema - model, average, arc, hellaswag, mmlu, truthfulqa and more, so you can filter, join and export without any glue code. No account or API key required to try it.

🎯 Target Audience	💡 Primary Use Cases
Analysts, developers, researchers and operators who need this data as clean rows	Data collection, enrichment, monitoring, and feeding apps, sheets or models

📋 What the Open LLM Leaderboard Scraper does

• Collects the records you ask for as clean, structured rows
• Returns 12 fields per record including model, average, arc, hellaswag, mmlu, truthfulqa
• Consistent, flat schema - ready for a spreadsheet, database or app
• Exports to CSV, Excel, JSON, or XML

💡 Why it matters: the raw source returns nested, paged responses. This Actor flattens them into one clean row per record, so the data drops straight into your workflow.

🎬 Full Demo (🚧 Coming soon)

⚙️ Input

Field	Type	Description
search	string	Filter models by case-insensitive substring match on the model name.
maxItems	integer	Cap on the number of results. Free: 10. Paid: up to 1,000,000.
precision	string	Filter models by reported precision.
minParams	integer	Filter models with at least this many parameters in billions.
maxParams	integer	Cap on the number of results. Free: 10. Paid: up to 1,000,000.
sortBy	string	Field used to sort the result set.

{"maxItems":25,"maxParams":25}

📊 Output

Each item is one record with 12 fields:

Field	Type	Description
model	string	Model
average	number	Average
arc	string	ARC
hellaswag	string	HellaSwag
mmlu	string	MMLU
truthfulqa	string	TruthfulQA
winogrande	string	Winogrande
gsm8k	string	GSM8K
params	number	Params (B)
precision	string	Precision
submitter	string	Submitter
scrapedAt	string	Scraped At

✨ Why choose this Actor

• Clean, flat schema - ready to use
• No API key or account needed
• Export to CSV, Excel, JSON, or XML
• Pay-per-event billing - only pay for records returned

🚀 How to use

1. Create a free account w/ $5 credit
2. Open the Actor on Apify Console
3. Fill in the input fields
4. Click Start
5. Download the results as CSV, Excel, JSON, or XML

💼 Use cases

Research and analysis - study the data at scale in your own tools.

Enrichment - add structured records to a database, sheet or app.

Monitoring - track changes over time with scheduled runs.

Datasets - build clean datasets for reporting, dashboards or models.

🔌 Automating Open LLM Leaderboard Scraper

Trigger the Actor via Make, Zapier, Slack, Airbyte, GitHub Actions, Google Drive, n8n or any HTTP webhook on Apify's integration platform.

🤖 Ask an AI assistant about this scraper

Paste this README into ChatGPT, Claude, Perplexity or Copilot and ask how to integrate the data into your workflow.

❓ Frequently Asked Questions

❓ Do I need an API key? No - just set the input and run.

❓ Free users? Capped at 10 items per run - perfect for testing.

❓ What output formats? CSV, Excel, JSON, or XML - pick on the Storage tab.

❓ Is the data live? Yes - fetched at run time from datasets-server.huggingface.co.

❓ Can I schedule it? Yes - use Apify's scheduler or any webhook integration.

🔌 Integrate with any app

Apify provides webhooks, REST API, scheduled runs, dataset downloads, and integrations with Make, Zapier, n8n, Slack, Google Drive, Airbyte, GitHub Actions, and more.

💡 Pro Tip: browse the complete ParseForge collection for more scrapers and all-in-one combos.

🆘 Need Help? Open our contact form

⚠️ Disclaimer: independent tool, not affiliated with any data source named above. Only publicly available data is collected.