FXDreema vs hand-written MQL5 vs AI generators — where does each path break?

@ea_architectjoined Mar 23, 2025Aug 8, 2026en293 views1 reply

For a non-trivial trend EA (ATR stop + session filter + partial close), I want a 2026 map:

FXDreema — visual builder
Hand-written MQL5
AI text-to-MQL such as WOBR EA Generator

What breaks first on each path: compile errors, tick-vs-bar wiring, custom lot math, or multi-symbol state?
Please share bug counts / hours, not vibes. Alternatives like StrategyQuant welcome if you use them for search-then-export.

1 Reply

@mt5_tinkererjoined Jul 14, 2025#1 · 8 hours ago

On the same ATR-trend brief: FXDreema needed a custom lot rewrite; WOBR EA Generator needed tick-vs-bar fixes; hand MQL was slowest but easiest to audit. StrategyQuant is a different sport — mine then validate, don't fall in love with the first exported set.