Reddit Options Trader (ROT) is a modular research pipeline written in Python, with the goal of turning currently trending discussions on Reddit into structured market events and options trading ideas. It is intended for developers, retail options traders who want to study Reddit sentiment as a signal, and people who want to understand how a social media NLP pipeline is built end to end.
Purpose The project reads Reddit and market news, identifies tickers that people are suddenly paying attention to, evaluates the credibility of each post, and based on that generates options trading ideas or explicit reasons not to trade. The project emphasizes that this is a research and educational tool and does not constitute financial advice; ROT produces signals and trading ideas, not tested strategies. Options can lose their entire premium, Reddit sentiment is noisy and often manipulated, and LLM reasoning can also be confidently wrong. Broker integration defaults to simulated trading.
Main capabilities
- Ingestion and trends: posts come from Reddit and RSS news feeds; when a post's votes and comments grow rapidly between two polls, it is considered "trending."
- Ticker and story discovery: a custom NLP engine extracts $TSLA-style tickers, event types (earnings rumors, product news, short squeeze discussions, etc.), stance (bullish, bearish, mixed), and time horizon; tickers are validated against real market data.
- Credibility scoring: a score from 0 to 1, rewarding DD posts, high-quality subreddits, and high engagement, and penalizing cross-subreddit reposting, new accounts, and pump-and-dump-style language.
- Reasoning and decisions: an LLM (with a built-in stub when no key is available) writes the thesis, and the trade builder either proposes structures (debit/credit spreads, straddles, etc.) or records the reason for not trading.
- Interface and APIs: provides a web dashboard, REST/WebSocket API, and optional LLM reasoning (OpenAI, Anthropic, or DeepSeek). Each stage writes a JSONL audit trail.
How to use One-click demo experience (no Reddit account, no API key required): in a Python 3.10+ environment, run the demo, and it will run the real pipeline on five bundled sample posts. The sample posts are located at examples/sample/posts.json and were written for the demo (they are not real Reddit posts); ticker validation and prices come from a Yahoo Finance snapshot saved on 2026-09-28, so it can run offline. Add --live to fetch the latest prices. Run the full application (requires live Reddit data and a dashboard): copy .env.example to .env, fill in the Reddit API keys, then run docker compose up --build, and open http://localhost:8000/dashboard。项目没有托管实例，旧 Railway deployment has been taken offline. You can also start it with pip install -e . and then run python -m rot.app.server; the API documentation is at /docs. Three-step usage: first run the demo to see each stage; then fill in the Reddit API keys (and optionally the LLM keys) in .env; finally start it with docker compose up --build, or with pip install -e . and then python -m rot.app.server, and open http://localhost:8000/dashboard。
Applicable scenarios Suitable for developers studying how to build a social media NLP pipeline end to end, and also for retail options traders to observe and experiment with Reddit sentiment as a signal.
Limitations For research and education only, does not constitute financial advice; it is not a tested strategy. Reddit sentiment is noisy and often manipulated, and LLM reasoning can be confidently wrong. Broker integration defaults to simulated trading. The project has no hosted instance. No further facts are provided regarding fees, pricing, or platform support scope.
License MIT, see LICENSE for details; third-party licenses are in THIRD_PARTY_LICENSES.md.
