LLM Benchmark Aggregator

What does Llm Benchmark Aggregator do?

Scrape LLM benchmark sites (MMLU, HumanEval, MATH). Aggregate scores across models for comparison tables. It runs on the Apify platform and delivers structured data in JSON, CSV, or Excel format, ready for analysis, integration, or automation workflows. Llm Benchmark Aggregator handles pagination, retries, and proxy rotation automatically so you can focus on using the data.

Why use Llm Benchmark Aggregator?

• No coding required — configure inputs in a simple web UI and click Start
• Export anywhere — download results as JSON, CSV, or Excel, or connect via API
• Scheduled runs — set up recurring scrapes to keep your data fresh (hourly, daily, weekly)
• Scalable — process hundreds or thousands of items with automatic proxy rotation and retry logic
• Integrations — connect to Google Sheets, Slack, Zapier, Make, webhooks, and more through the Apify platform

How to use Llm Benchmark Aggregator

1. Navigate to the Llm Benchmark Aggregator page on Apify Store and click Try for free
2. Configure your input parameters (see Input Configuration below)
3. Click Start and wait for the run to complete
4. View results in the Output tab — use the formatted table or switch to raw JSON
5. Download your data as JSON, CSV, or Excel, or access it via the Apify API

Input configuration

Field	Type	Description	Default
Benchmark URLs	array	List of LLM benchmark pages to scrape	['https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard']
Max Results	integer	Maximum number of results	100

Output data

The actor stores results in a dataset. Each item in the dataset represents one extracted record with structured fields. You can preview the data in the Output tab's formatted table view.

Key output fields include: URL, Score, Rank, Scraped At, Benchmark.

Example output:

{
  "url": "https://example.com/url",
  "score": 4.5,
  "rank": "Example Rank",
  "scrapedAt": "Example Scraped At",
  "benchmark": "Example Benchmark",
  "_fallback": "Example  Fallback"
}

Each run also produces an execution log with detailed information about pages processed, items extracted, and any errors encountered.

Cost of usage

Llm Benchmark Aggregator uses Pay-Per-Event pricing (Mid tier). Each successfully extracted result costs approximately $0.0008 ($0.75 per 1,000 results).

On a free Apify plan ($5/month platform credit), you can extract approximately 6,666 results per month.

Example: Extracting 1,000 results would cost approximately $0.75.

Tips and advanced usage

• Proxy configuration: This actor uses lightweight HTTP requests for fast, efficient scraping. For sites with rate limiting, the actor automatically rotates proxies.
• Large datasets: For runs with thousands of results, increase the memory allocation in Run Options to speed up processing. The actor automatically manages request queues and pagination.
• Scheduled runs: Use Apify Schedules to run this actor on a recurring basis. Combined with integrations (webhooks, Google Sheets, Slack), you can build automated data pipelines that keep your datasets up to date.
• API access: Every dataset is accessible via the Apify API. Use the REST API or official Python/JavaScript clients to integrate results directly into your applications.

Useful Links

• LMSYS Chatbot Arena
• Hugging Face Open LLM Leaderboard
• Apify Proxy

Related Actors:

• Groq Model Scraper

• Ollama Model Scraper

• Ai Pricing Comparator

• More actors by consummate_mandala

• GitHub