本页保存的是公开项目资料快照，阅读过程不需要连接 GitHub。

GammaGrid — Open-Source Options Gamma Exposure (GEX) & Positioning Dashboard

Track dealer gamma exposure (GEX), max pain, open interest, and the IV surface for your whole options watchlist — not just SPY. Self-hosted, open source, built on free market data.

图片：License: AGPL v3 图片：Runs on Docker

No coding required. If you can install an app and copy-paste one command into a terminal, you can run GammaGrid. No Python, no config files, no programming experience needed — see Quick start below.

图片：GEX Heatmap: strike × expiry gamma exposure matrix, with Call Wall, Put Wall, Gamma Flip, and Replay

Quick start (no coding required)

1. Install Docker Desktop — docker.com/products/docker-desktop. It’s free; just click through the installer like any other app.
2. Download this project — click the green Code button at the top of this page → Download ZIP, then unzip it. (Comfortable with git instead? git clone this repo.)
3. Open a terminal in the unzipped folder — on Mac: right-click the folder → New Terminal at Folder. On Windows: open the folder in File Explorer, type cmd in the address bar, press Enter.
4. Run one command:

   docker compose up

   (Older Docker installs: use docker-compose up instead — same effect.)
5. Open http://localhost:8501 in your browser. That’s it — GammaGrid is running.

Data is saved to data/options.db on your machine and survives restarts. Press Ctrl+C in the terminal to stop the app; run the same command again to bring it back up with your data intact.

What you get

• Dealer gamma exposure (GEX) — per-expiry profile and a strike × expiry heatmap, with Call Wall / Put Wall, Gamma Flip level, and historical Replay
• Max Pain for any expiry
• Open interest, including day-over-day OI Delta sorted by the size of the move
• IV surface (3D volatility surface) plus per-expiry skew and volume-weighted average IV over time
• Options screener with the full set of greeks (delta, gamma, theta, vega, rho, vanna, charm) and range filters — not just delta/IV like most free tools
• Unusual activity detection — flags contracts whose volume is a statistical outlier against that specific contract’s own history, not a flat threshold
• Put/Call Ratio and per-contract price/IV/greeks history with pinning
• Works for any ticker with a listed options chain — build your own watchlist, not a single fixed symbol

Usage

1. In the left sidebar, enter a ticker (e.g. AAPL) and click Add — it appears in the watchlist.
2. Click Collect data — the app fetches the current option chain for every watchlist ticker via Yahoo Finance and saves a snapshot. Collection is manual; there is no automatic schedule.
3. Pick a ticker in the dropdown above the tabs to open the metrics:
   • Overview — Put/Call Ratio over time and the IV surface
   • Max Pain / GEX — max pain and the approximate gamma-exposure profile for a selected expiry
   • GEX Heatmap — strike × expiry GEX matrix with Call/Put Walls, Gamma Flip, and snapshot Replay
   • Volatility (IV) — ticker-average IV over time and a chain skew slice
   • Contract — price, IV, and greeks history for a specific contract, with pinning
   • Screener — every contract of the latest snapshot with greeks and range filters
   • Unusual Activity — contracts with anomalous volume in the latest snapshot
   • OI Delta — open interest change between the two latest calendar days

Most history-based metrics (other than Put/Call Ratio, average IV, and OI Delta) need several days of collection — some charts require at least two snapshots. Click Collect data daily (or a few times a day) to build up history.

Screenshots

IV surface	Options screener
图片：3D implied volatility surface across the option chain	图片：Options screener with the full set of greeks and range filters

All screenshots above are real GammaGrid output — SPY/QQQ/MSFT via a live collection, no mockups.

FAQ

What is dealer gamma exposure (GEX)? It’s an estimate of how much options market makers are net long or short gamma across a ticker’s option chain. Positive GEX suggests dealer hedging tends to dampen price moves; negative GEX suggests it can amplify them. GammaGrid computes this via Black-Scholes as an approximation from options open interest — it is not a measure of actual market-maker positions, which aren’t public data (see the disclaimer below).

Is this a real-time options flow scanner? No — GammaGrid takes periodic snapshots of the option chain (on-demand, via the Collect data button), it does not stream live trade-by-trade tape. If you need tick-by-tick sweep/block alerts, that’s a different category of tool. GammaGrid is for tracking positioning and structure (GEX, max pain, OI, IV) across a watchlist over time.

For developers: running from source

python -m venv .venv && source .venv/bin/activate
pip install -r requirements.txt
PYTHONPATH=. streamlit run app/dashboard.py

PYTHONPATH=. is required: Streamlit adds the script’s own directory (app/) to sys.path, not the project root, and without it the app fails with ModuleNotFoundError: No module named 'app' — the imports in dashboard.py (from app import ...) expect the project root to be visible in sys.path. In Docker the same thing is handled by ENV PYTHONPATH=/app in the Dockerfile.

Data source limitations

yfinance is an unofficial wrapper around Yahoo Finance, with no SLA or official support. Expect possible data delays (15–20 minutes), irregular intraday open-interest updates, and temporary blocks under frequent requests. The app logs collection failures (visible on the dashboard after clicking Collect data) but makes no attempt to circumvent blocks.

Want it hosted, with zero setup?

A hosted version of GammaGrid (no Docker, no local install) is planned. Join the list at gammagrid.io to hear when it’s ready.

Get involved

Questions, feedback, or found a bug? Email hello@gammagrid.io or open an issue. If GammaGrid is useful to you, starring the repo genuinely helps — it’s the main signal used to decide what gets built next.

License

AGPL-3.0.

Disclaimer

This software is for informational and educational purposes only and does not constitute investment advice. All metrics are approximations built on delayed, unofficial data.