AI Experts

Scalperology AI targets short, frequent moves on gold during the high-liquidity overlap of the London and New York sessions. The neural network filters out low-quality entries that pass the rule-based filters alone, materially reducing the trade count but improving the win rate of the trades that are taken.

Key numbers: - Pair: XAUUSD - Timeframe: M1 - Min deposit: $500 - Broker style: ECN / raw spread, XAUUSD spread under 1.5 pips - VPS: strongly recommended

See the dedicated product page for live statistics and the latest review.

Breakopedia identifies the prior session's high-low range and arms breakout orders just before the next session opens. The confirmation filter — a combination of momentum and order-flow proxies — rejects breakouts that show characteristic false-break behaviour (weak volume, immediate reversal).

Because Breakopedia trades 1–3 times per day per pair, it is much easier to evaluate on demo and far more tolerant of broker execution quirks than a scalper. It also produces a flatter equity curve, which often suits prop-firm daily-drawdown rules better than scalpers do.

Trend-followers profit from sustained directional moves and lose money in range-bound markets. Trendopedia's regime filter is what differentiates it from naive trend EAs: when the filter classifies the regime as ranging, the EA stops opening new trades and either trails the stop on existing positions or closes them, depending on the preset.

The trade frequency is low (typical 5–15 trades/pair/month), so a meaningful live evaluation takes 6–12 months. Trendopedia is the easiest EA in our line to forget about — it does not require a VPS for monthly monitoring, although a VPS is still recommended for execution consistency.

Where Scalperology is a rules-based scalper with an ML overlay, GoldStrike is ML-driven end-to-end: the entry decision, the exit timing, and the position-sizing scale all come from a model that is retrained weekly on the latest market data.

The higher price reflects (a) the ongoing model-retraining infrastructure, (b) priority support, and (c) a 30-day trial period. GoldStrike is not the right choice for a $500 account — the minimum deposit is positioned around the level at which the ML's risk-control assumptions hold up.

See /experts/goldstrike-ai for the full review and live statistics.

Our four flagship EAs are the only ones we recommend without qualification because they are the only ones we control end-to-end. We choose the broker, we publish the account, we retrain the model, we ship updates.

Third-party EAs in the /experts catalog go through our six-point methodology (see /about/methodology) but the quality range is enormous. Some are excellent and we praise them. Many are mediocre. A few are outright scams that we flag in the reviews. Always read the full review, not just the rating.

Martingale-style position sizing produces an attractive equity curve in normal market conditions and a catastrophic drawdown in abnormal ones. The math is unforgiving: a martingale that doubles after every loss requires exponentially growing capital to survive a losing streak that is statistically certain to eventually occur.

Grid systems share the same failure mode in disguise. Both classes regularly appear in EA marketplaces with backtest equity curves that look like straight lines — until you check the open positions during a strong directional move and find the account is one stop-out away from zero.

The EAs we sell use fixed risk-per-trade sizing or scale by recent volatility, never by recent losses.

Standard input parameters cover 90% of customisation needs: lot sizing, risk %, allowed pairs, allowed hours, news filter on/off, magic number, stop-loss override.

Core strategy modifications — for example, replacing the entry filter with your own — require the source code (`.mq5`) and a custom-development engagement. We accept bespoke MQL5 work for serious requests; pricing depends on scope. Email [email protected] or contact us via Telegram.