Priced like a tool, not a terminal.

One person, one research stack. Every tier runs the same engine — the difference is depth, not quality.

Free

$0 forever

Enough to find out whether the idea survives contact with data.

most useful

$19 per month

The full archive and the ML hooks. For a strategy that earned a budget.

$49 per month

Team seats and programmatic access for small desks.

Billing is a placeholder in this build — plan switches apply instantly in Settings so every gate is testable.

Is the market data real?: Out of the box, no — the platform ships with a deterministic synthetic feed (16 years of daily bars, regime shifts, a shared market factor) so every feature works offline. Add a Polygon, Alpaca, or Yahoo Finance key in Settings and the same engine runs on real bars.
Does paper trading use real money?: Never. Paper sessions trade a virtual book against the live feed. The order path is built so a broker paper API (e.g. Alpaca paper) can be wired in without touching the dashboard.
What does the ML signal actually do?: It builds a standard feature block (21/63-day momentum, 20-day z-score, realized vol) and scores it with your model — either an HTTP inference endpoint or a local sklearn/XGBoost pickle. Without a model it falls back to a labeled linear blend so runs never break.
How is slippage modeled?: Three components per order: a fixed per-share cost, a proportional basis-point cost, and square-root market impact scaled by participation in daily volume. All three are tunable per strategy version and itemized in results.
Can I cancel?: Yes, anytime. Strategies and results stay readable on the Free tier; only the gates (strategy count, history depth, ML) re-apply.