AI Models & LLM Benchmark Tracker

Track, compare, and extract frontier & open-source LLM benchmark metrics across LMSYS Chatbot Arena, Open LLM Leaderboard, and coding benchmarks in structured JSON/CSV format.

The AI Models & LLM Benchmark Tracker provides standardized intelligence on artificial intelligence models, including Arena Elo ratings, MMLU general reasoning accuracy, HumanEval/coding benchmarks, MATH problem solving, license types, context windows, and parameter sizes.

---

⚡ What Does This Actor Do?

• Consolidated Leaderboard Intelligence: Aggregates metrics from leading evaluation sources into a unified, queryable dataset.
• Granular Filtering: Filter models by minimum Arena Elo rating (e.g. models rated 1200+ Elo), license type (open-source vs proprietary), and specific benchmark suites.
• Pay-per-Event Pricing: Pay only for results produced ($0.002 per model record), with $0 startup fees.
• AI Agent & MCP Native: Fully compatible with Model Context Protocol (MCP) clients, Cursor, Claude Code, and autonomous AI data agents.
• Multiple Export Formats: Download data instantly as JSON, CSV, Excel, XML, or access programmatically via REST API.

---

🛠️ Input Configuration

Field	Type	Default	Description
benchmarks	Array	["arena-elo", "open-llm", "coding-benchmark"]	Select evaluation suites to track.
max_models	Integer	50	Maximum number of models to return (1 - 200).
min_arena_elo	Integer	1100	Minimum Chatbot Arena Elo threshold.
license_filter	String	"all"	Filter by license: all, open-source, or proprietary.

---

📊 Sample Output

{
  "model_name": "Claude 3.5 Sonnet",
  "organization": "Anthropic",
  "arena_elo": 1285,
  "coding_score": 93.7,
  "mmlu_score": 88.7,
  "math_score": 78.3,
  "license": "proprietary",
  "parameters_billion": null,
  "context_window": 200000,
  "data_source": "LMSYS Chatbot Arena / Open LLM Leaderboard Consolidation",
  "scraped_at": "2026-09-04T02:00:00Z"
}

---

💰 Pricing

This Actor utilizes Apify's Pay-Per-Event (PPE) pricing:

• Startup Cost: $0.00
• Per Result: $0.002 USD ($2.00 per 1,000 model benchmark rows)
• Apify Store Discounts: Enabled for paying platform tiers (Personal, Team, Enterprise).

---

🚀 How to Run via API or CLI

Python

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("papa_developers/ai-model-benchmarks-scraper").call(run_input={
    "max_models": 25,
    "min_arena_elo": 1200,
    "license_filter": "open-source"
})

for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["model_name"], item["arena_elo"])

Apify CLI

$apify call papa_developers/ai-model-benchmarks-scraper -i '{"min_arena_elo": 1200}'