CounterQuant CS2 Demos

The largest open collection of professional CS2 demo files (.dem)

Raw competitive Counter-Strike 2 demos collected from HLTV.org — focused on Tier 1 & Tier 2 matches.

Curated and maintained by Eimantas Kulbe (KEDevO) as part of the CounterQuant esports analytics project.

📊 Dataset Overview

• Total Size: 3.74 TB+ (and continuously growing)
• Time Period: January 2024 – Present
• Focus: Tier 1 & Tier 2 professional matches
• Update Frequency: New demos added within 24–48 hours after matches
• Last Updated: May 20, 2026

🎯 Goal & Philosophy

To provide the global research, analytics, and esports community with unrestricted access to high-quality professional CS2 demo files.

Philosophy: Raw data only. No gatekeeping. Parse it, analyze it, and build whatever you want.

📁 File Structure

data/
├── 2024/
│   ├── tier1/
│   └── tier2/
├── 2025/
│   ├── tier1/
│   └── tier2/
└── 2026/
    ├── tier1/
    └── tier2/

Example filepath: data/2025/tier1/2401337/vitality-vs-g2-inferno.dem

🚀 Quick Start

Download specific year + tier (Recommended)

from huggingface_hub import snapshot_download

# Example: Download all 2025 Tier 1 demos
snapshot_download(
    repo_id="KEDevO/CounterQuant-CS2-Demos",
    repo_type="dataset",
    allow_patterns="data/2025/tier1/**",
    local_dir="./cs2_demos",
    resume_download=True
)

Parse a demo

from demoparser2 import DemoParser

parser = DemoParser("data/2025/tier1/2401337/vitality-vs-g2-inferno.dem")

kills = parser.parse_event("player_death")
rounds = parser.parse_ticks(["total_rounds_played", "cash_spent_t", "cash_spent_ct"])

What You Can Build

• Advanced win probability & clutch models
• Player chemistry & synergy analysis
• Utility usage and trade efficiency metrics
• New generation rating systems
• Tactical & game-theory research
• ML models for movement, positioning, and decision making

Related Projects

• CounterQuant Bronze — Structured & parsed version (Parquet/JSON)
• CounterQuant Platform — Live analytics dashboard & API

Citation

@dataset{kulbe2026counterquant,
  author       = {Eimantas Kulbe},
  title        = {CounterQuant CS2 Demos},
  year         = {2026},
  publisher    = {Hugging Face},
  url          = {https://huggingface.co/datasets/KEDevO/CounterQuant-CS2-Demos},
  note         = {Continuously updated professional CS2 demo collection}
}

License

This dataset is licensed under CC BY 4.0 — you are free to use, modify, and build upon it with proper attribution.