Aurum AI is a free MT4 robot for trading gold (XAUUSD). It opens trades only in the direction of the prevailing trend, places a fixed Stop Loss and Take Profit on every position, and stays out of martingale, grid, and locking. The original MQL5 signal that shipped with this article in December 2024 has been removed, so we are republishing this review against a fresh April-May 2026 verified signal on an IC Markets Raw Spread account.
This update focuses on what the new signal actually tells us about the EA in 2026, where the strategy is honest, and where the limits of the current track record start to bite.
TL;DR – Free MT4 XAUUSD trend EA. Hard stops, no martingale, no grid. New live signal is real but tiny, so the headline win rate is closer to “no losers yet” than to “proven edge”. Overall rating: 3.5 / 5.
Core Strategy and Logic
Aurum AI is a one position at a time XAUUSD robot. It looks for setups in the direction of what it reads as the dominant trend on the lower timeframes and closes inside its fixed Stop Loss or Take Profit levels.
- Trend-aligned entries – The EA does not stack counter-trend positions, so it avoids the typical “fade the move” trap that kills most aggressive gold bots.
- Hard Stop Loss – Every order is placed with a fixed Stop Loss in points (the largest loss the trade can take before being closed automatically). No floating loss is ever held open hoping for a recovery.
- Fixed Take Profit – Each trade carries a fixed Take Profit. There is no break-even shuffle, no doubling, and no grid layering.
- No locking or hedging – Two opposing positions on the same instrument are never opened to mask drawdown.
- Auto-risk or manual lot – Users either set a fixed lot size or enable an automatic risk-percentage calculator. The auto mode is what makes the EA usable on prop-firm accounts.
This design is structurally close to the trend following XAUUSD robot we reviewed earlier this year and clearly different from grid-based gold systems like XAUUSD Trend Grid, which carries multiple open trades by design.
Key Features of Aurum AI EA
- Platform – MetaTrader 4 (MT4). The download distributed here is the MT4 build only; the vendor’s public MQL5 signal happens to run an MT5 deployment of the same logic.
- Symbol – XAUUSD (gold) only
- Timeframe – M5 attachment
- Strategy – Trend-following with hard SL and TP, no recovery methods
- Risk control – Auto-lot from a user set risk percentage, or a manual lot override
- Prop firm friendly – Single position and fixed stops fit FTMO, Moneta Funded, and MyForexFunds rule sets
- Backtest coverage – Vendor states the EA was tuned and tested on data from 2012 onwards
Please test in a demo account for at least a week first. Then, please familiarize yourself with and understand how this Aurum AI Robot works and only use it on a real account.
Recommendations for Aurum EA
- Minimum account balance of 100$.
- Works best on GOLD. (Work on any Pair)
- Work best on M5 TimeFrame. (Work on any TimeFrame)
- Aurum EA should work on VPS without interruption and low latency to reach stable results. So we recommend running Aurum AI EA on a reliable VPS (Reliable and Trusted FOREX VPS – FXVM)
- Low Spread account is also Recommended (Find the Perfect Broker For You Here)
If you plan to use auto-risk mode, set the percentage so the dollar risk per trade matches your prop-firm daily loss limit. Our position size calculator makes that conversion fast.


Performance Review: Live Data Analysis




Why the headline numbers flatter the EA
A 100% win rate on a clean equity ladder reads well, but it lands on a sample small enough that statistically, the EA has not yet earned the benefit of the doubt. Every fixed-TP system goes through stretches like this before the first stop loss fires. The honest read is “no losers yet” rather than “proven edge”. What we are watching for in the next two to three months is how the first losing trade behaves, because that is the trade that tells us whether the published Stop Loss is real or whether the EA quietly drags it.
What the equity curve does prove
Two things are credible even at this sample size. The risk profile matches the marketing: drawdown stayed well inside any prop-firm daily limit, deposit load never approached margin pressure, and there is no sign of a hidden lock or grid layer waiting to unwind. The trade distribution is also balanced between long and short entries, which rules out a system that only works because gold has been trending one way recently. Both of those are structural facts the screenshot supports independently of the win rate.
What it does not prove
Trade frequency is the real friction. Roughly one entry per week, with the latest trade two weeks before our capture, means an account using Aurum AI as its only EA will spend most of its time waiting. That is fine for a portfolio slot, less fine for a trader expecting daily activity. It also means a $150 balance grows in chunks, not smoothly, and one losing trade at default risk will hand back a meaningful share of the published gain.
What we would track next
Run the EA on a $100-$200 demo for at least a month before going live, and look specifically at the first three losing trades. If the actual loss matches the fixed Stop Loss the inputs show, the system behaves the way the vendor advertises. If a “Stop Loss” trade closes for a smaller-than-expected amount, that is the warning sign that something other than the published rule set is in play. Comparing the live signal against a higher-frequency gold scalping EA or a fixed-stop gold prop-firm robot over the same window is the cleanest way to see whether Aurum AI is genuinely competitive or just early in a quiet streak.




Aurum AI EA

Free MT4 expert advisor for XAUUSD that trades the dominant trend with fixed SL and TP, no martingale or grid, suitable for prop-firm risk rules.
✓ Pros
- Hard Stop Loss and Take Profit on every trade
- No martingale, grid, or locking
- Single open position at a time keeps margin usage modest
- Compatible with prop-firm daily and overall loss caps
✗ Cons
- Public live signal is only two months old as of May 2026
- Original 2024 signal was removed, breaking long-term continuity
- Trade frequency is low, often a week or more between entries
- "AI" branding is marketing language, not a verifiable model
Summary
Aurum AI is one of the more honestly designed free gold EAs on the market in 2026. The risk model matches what the vendor advertises and the equity curve has no hidden recovery games behind it. The catch is that the public live signal has been running for only two months, and that is too short for the win rate to mean what it looks like at first glance.
Frequently Asked Questions
Conclusion
While Aurum AI presents itself as an innovative tool for gold trading with AI integration, traders should critically evaluate its actual capabilities. Like many EAs in the market, artificial intelligence claims should be viewed with skepticism unless proven through consistent results. For those willing to experiment, Aurum EA could be another option to explore within the growing ecosystem of trading tools. Always remember to approach trading with prudence and use proper risk management to safeguard your investments.

no web request ???
Back test curve looks like as if this EA is another history reader.
Admin, when working on the source code, you can certainly identify this aspect and confirm or infirm it. Could you please do that systematically ? Thank you.
Here a typical back test curve end of data readers, when no more historical data are available.
Grazie mille Eric , avevo notato anche io la stessa cosa , cosi’ i BT sono inutili in effetti.
Indeed Marco and by the way, I’m still wondering why the Admin posts this kind of EAs. He has access to the code so he should be aware of the cheat !
Very sketchy. I tried to backtest with data before 2012, and no trade opened. It is curious that on the 11th of January 2012, the first trade started and was win after win. It seems like it is programmed to deliver only wins, ignoring the losers. It probably is like Eric says, and those last losses from December 2024 will be eliminated in the next update…
NO, i tried backtest in 2023 and it was not good.
in 2024 everything is good. because the base is EMA and this strategy for 2023 is losses because in 2023, the market doesn’t have volatility. I did copy this strategy and for me was the same as this.
the strategy is good for 2024 and in 2025 if you have a chance volatility be the same as in 2024
is this EA still working? I tried it but it didnt work at all