Is X Fusion AI legit?

Unknown. EALegit has not audited X Fusion AI. A listing is not a verdict that it is legit.

Has EALegit independently reviewed X Fusion AI?

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

Does X Fusion AI 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 $179

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: $179

Submit source if you own this EA View on MQL5 ↗

X Fusion AI3.2

Developed byChen Jia Qi5 EAs on record · MQL5 rating 3.6

View on MQL5 ↗

Not testedAI / neuralTrend followingBreakoutGBPUSDM15MetaTrader 5

Seller claims

grid no · martingale no · SL yes

EALegit summary — unaudited listing

X Fusion AI is a MetaTrader expert advisor for automated forex trading on GBPUSD, recommended on the M15 timeframe with a minimum deposit of around $500. It combines trend, momentum, pullback, and breakout-style detection rules with a neural-like adaptive engine that adjusts internal component weights to market conditions. Version 3.0 replaced the previous martingale-style scaling in Strategy 2 with a one-order-at-a-time structure, which is disabled by default. Risk management includes dynamic stop loss and take profit levels, break-even and trailing logic, losing-sequence protection, volatility-based activity reduction, and news-time filtering.

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.