Unknown. EALegit has not audited Mad Turtle. A listing is not a verdict that it is legit.

Has EALegit independently reviewed Mad Turtle?

Not yet. Listing is seller materials. The owner can submit source from For developers.

Does Mad Turtle use grid or martingale?

Seller materials do not claim grid or martingale. Not audited — we have not confirmed the code.

Can I trust the seller backtest?

No usable seller baseline is published and we have no lab replay. Treat the seller backtest as untested marketing.

Unaudited seller listing — lab verdict pending

Untrusted tick model

This listing is built from seller materials. Lab replay not yet executed. Lab, audit, live tracking and verified reviews stay empty until we compile and run real-tick replay.

Submit source if you own this EA

Seller price $1599

Curve replay

Backtest vs lab reality

Zero corroboration

Seller marketing curve (claimed)Lab replay not run

Execution engine locked

Real-tick replay has not run for this listing.

Claimed profit factor

— (marketing)

Real-tick PF

Pending

Seller price: $1599

Submit source if you own this EA View on MQL5 ↗

Mad Turtle10.1

Developed byGennady Sergienko2 EAs on record · MQL5 rating 4.44

View on MQL5 ↗

Not testedAI / neuralTrend followingXAUUSDH1 / M15 / H4MetaTrader 5

Seller claims

grid no · martingale no · SL yes

EALegit summary — unaudited listing

Mad Turtle is a fully automatic MetaTrader 5 expert advisor for trading gold (XAUUSD) on the M15-H1 timeframes. It uses machine learning models deployed locally via MT5's ONNX library, with ensemble and meta-model filtering, holding single positions from several hours to several days rather than scalping. The seller states it avoids grid and martingale methods and uses stop losses as an integral part of the strategy, without trailing stops or early profit-taking. It requires a minimum deposit of about 500 USD and is stated to run without pre-configuration on any broker.

Lab results

Canonical test windows — a slot shows a report only after the run finishes and passes checks.
2024.01.01 → 2026.09.01	failed — see provenance
2024.01.01 → 2025.01.01	failed — see provenance
2025.03.01 → 2025.07.01	failed — see provenance
2026.01.01 → 2026.07.01	failed — see provenance

Seller evidence vs lab result

No usable seller backtest on file. The seller listing does not publish a testable baseline (dates, settings and metrics together), so reproduction is blocked — not failed. Lab results above stand on their own.

Seller evidence is transcribed from dated sources and reviewed before publication. It never changes the performance score or the audit verdict.