Plain English in. Backtested strategy out.

An open sandbox of Tradevo's strategy engine. Your description is compiled by an LLM into a constrained spec — a fixed vocabulary of indicators, entries, exits, and risk rules; it can't emit arbitrary code — then run through a deterministic backtest over the last 180 days of real daily data, net of modeled swap costs. No signup, nothing saved.

~15 seconds · free · rate-limited

What this is — and isn't

This sandbox is an overfitting machine by construction. Iterate on prompts long enough and you'll find a curve that looks great in-sample. That's a property of backtests, not a discovery.
A good-looking result here earns nothing in the real product: publishing to Tradevo's marketplace requires a historical replay window, a numeric evaluation, and human review. Most candidates fail.
Costs are modeled per-swap; live execution is the arbiter. The LLM only writes the spec — the backtest itself is deterministic engine code with no AI in the loop.
Past performance does not predict future results. Nothing here is financial advice.

Full methodology: how strategies are validated and tracked.

The live product runs curated strategies 24/7 from a wallet only you control — app.tradevo.co · US residents only