2.7 Installing, Running & Expected Performance

2.7.1 Requirements

From the README:

• Python 3.8+

• requirements.txt dependencies (aiohttp, etc.).

• A stable internet connection.

• At least a few GB of RAM; nothing insane.

2.7.2 Quickstart

1. Install dependencies:

   pip install -r requirements.txt

2. Run the engine:

   python spreadnet.py

3. Watch the terminal:

   Capture OS - Real-time Solana Arbitrage Detection
   ============================================================
   Monitoring DEXs: Jupiter, Raydium, Orca, Phoenix...
   Minimum profit threshold: 0.1%
   Press Ctrl+C to stop
   ...
   Stats: 20 opportunities | Avg: 0.75% | Best: 2.30%
   ``` :contentReference[oaicite:32]{index=32}  

2.7.3 Expected Performance

Indicative numbers (obviously market-dependent):

• Detection speed: < 1s from price change to detection.

• Opportunities/hour: ~20–100.

• Average detected profit: ~0.3–1.5% per opportunity.

• False positive rate: < 10% with default filters.

Market conditions impact:

• High volatility → more opportunities, wider spreads.

• Low volatility → fewer opportunities, thinner margins.

• Network congestion → slower execution, higher costs; detection still runs but your fills may suffer.

---

Bottom line: Capture Engine is the OSS foundation of Capture OS — a small, auditable Python engine that tells you, in plain numbers, where Solana markets are mispriced right now. Everything else (UI, bots, risk layers) is built on top of this.