Best Prediction Market Botsranked & reviewed · 2026 Open Kairos

Multi-strategybotfree

Polymarket BTC 15-Minute Trading Bot: what it does, how it decides, where it stops

Last checked 2026-09-08

VenuesPolymarket
PricingOpen source.
Typebot
GitHub stars583

Polymarket-BTC-15-Minute-Trading-Bot is a Python bot for Polymarket's 15-minute BTC up/down markets. It fuses three signal processors (spike detection, sentiment, price divergence) through weighted voting, sizes at $1 per trade with a 30% stop-loss and 20% take-profit, and exports metrics to Grafana.

It is built on NautilusTrader and uses Redis to switch between simulation and live without restarting. The README's own note says that switch is not stable yet.

What it does

Seven phases in the README: data sources (Binance WebSocket, Coinbase REST, Fear and Greed plus social sentiment, an optional Solana RPC), ingestion with validation and rate limiting, a NautilusTrader core, signal processors, a fusion engine, execution with a risk engine, and monitoring with Prometheus and Grafana. A seventh learning phase is a placeholder file.

Paper trades are viewable with view_paper_trades.py. Test mode trades every minute for debugging; normal mode every 15 minutes.

  • Env: POLYMARKET_PK, POLYMARKET_API_KEY, POLYMARKET_API_SECRET, POLYMARKET_PASSPHRASE, Redis host and port
  • Defaults: MAX_POSITION_SIZE 1.0, STOP_LOSS_PCT 0.30, TAKE_PROFIT_PCT 0.20, SPIKE_THRESHOLD 0.15, DIVERGENCE_THRESHOLD 0.05
  • python run_bot.py with the test-mode flag; python 15m_bot_runner.py with the live flag

Strategy

Spike detection, sentiment and divergence each vote; the fusion engine weights the votes and the risk engine caps the result at $1. The README says signal weights self-optimize from performance, and the FAQ cites roughly 75% win rate in early simulation runs while saying past performance does not guarantee future results.

No backtest results, no live results, no fee model are published.

Setup and requirements

Python 3.14 or newer per the README, Redis, and Polymarket API credentials plus a private key. The quick-start clone URL in the README points at a placeholder yourusername path rather than the real repo.

The repo contains patch_gamma_markets.py and patch_market_orders.py, described as temporary patches for the Polymarket API.

Activity and maintenance

583 stars and 172 forks as of 2026-09-08. Created 2026-02-15, last push 2026-02-28, 8 open issues. Python. Six months without a push; the GitHub API reports no license file despite the README's MIT badge. Treat it as unmaintained.

Community links go to a Discord and to a Telegram handle.

Where it falls short

A $1 cap makes it a learning tool, and the README says so. The mode switch is flagged unstable, the learning engine is a placeholder, and the API patches suggest the Polymarket client was already drifting when work stopped.

Single venue, single market type.

Who should use it

A developer who wants to see NautilusTrader wired to Polymarket and a multi-signal fusion layout with monitoring, and will bring the Polymarket client up to date.

Not for anyone who wants a maintained 15-minute BTC bot.

Questions about Polymarket BTC 15-Minute Trading Bot

Is the Polymarket BTC 15-Minute Trading Bot profitable?

The README cites roughly 75% win rate in early simulation runs and adds that past performance does not guarantee future results. No live results are published.

Is it free?

Yes. The README carries an MIT badge, though the repo has no LICENSE file.

Is it maintained?

No. Last push 2026-02-28 as of 2026-09-08.

What do I need to run it?

Python 3.14+, Redis, Polymarket API credentials and a private key. It caps each trade at $1.

signalsizefilledge

Bots die in execution. Terminals are execution.

Wanted a bot for speed and cross-venue vision? The free Kairos terminal trades Kalshi, Polymarket and Predict.fun from one book, no keys to manage.