SmartChoise is a Gold trading EA built around neural-network entries, adjustable position sizing and recovery management. The MT5 and MT4 releases share that broad idea, but their documented timeframes and controls differ. This review compares the two editions and examines what their settings, release history and public customer feedback can tell a buyer.
Our assessment: the extensive controls appeal to traders who want to configure and monitor a recovery system. The main consideration is what happens after an entry goes wrong: additional positions can increase exposure, and a convincing backtest does not establish dependable live performance. This is a source-based assessment, not an independent trading test. CheaperForex offers the products discussed.
Sources checked September 29, 2026. Version numbers below refer to the public MQL5 listings checked on that date.
MT4 and MT5: what actually differs?
SmartChoise MT5
The MT5 listing showed version 9.7, updated September 13, 2026, and identifies XAUUSD on M1. Its separate timeframe guide is broader: it discusses M1–M15, highlights M10 as a balance between activity and noise, and suggests beginning with demo observation on M1–M10. These are developer recommendations, not independently verified rankings of the timeframes. MT5 source listing. Developer timeframe guide.
MT5 has a substantial evolving control set. Its May 2026 release notes introduced first-entry confirmation using a fixed or ATR-based distance, a timeout, and momentum or pullback confirmation. Recovery trades remain governed by their own logic. The April update added a time-and-profit/loss exit check; September’s update refreshed the neural-network checkpoint and filtering. These changes are reasons to record the build and settings used in a test. MT5 release notes.
SmartChoise MT4
The MT4 listing showed version 1.0 and specifies XAUUSD on M15. It describes a simplified branch with fewer lot-increase options and an enhanced Trend strategy. It also says take profit or trailing stop replaces the internal profit-closing logic. That difference matters more than a shared product name: a setup copied across platforms may produce different exits. MT4 source listing.
There is an inconsistency in the MT4 description: an older feature bullet mentions support/resistance and candlestick detection, while the later explanation says only the enhanced Trend strategy was retained alongside the neural network. We have not assumed those optional strategies are available in MT4. The shared manual explains concepts, but the inputs in the actual MT4 build take precedence over later MT5 features.
Recovery is the central decision
The public manual describes adding trades in the same direction as the original position, with adjusted volume, and looking for a combined break-even or profit exit. The developer distinguishes this from martingale. That label alone does not answer the practical risk question: how large can the combined position become before the system closes it? Recovery explanation.
The maximum recovery-trade setting applies per chart. Running several instances can multiply the account’s total exposure, even when each chart has a limit. A maximum count also does not force an immediate recovery entry; the documentation says market conditions can delay it. A basket can therefore remain in drawdown without another trade appearing. Maximum recovery trades.
The MT5 manual also documents a recovery lot multiplier. Increasing it enlarges added positions; reducing it changes the pace of attempted recovery. Fixed lots separately override adaptive sizing. These are substantive strategy changes, not harmless cosmetic adjustments. A risk label such as Low cannot be assessed without the volume overrides, number of instances and active loss limits. Recovery volume control. Fixed-lot behavior.
Stops and filters need a clear scope
The manual distinguishes the MT5 Hard Stop for an EA instance from the setting that measures account-wide floating drawdown. For the latter, the documented action is to close the EA’s positions; it should not be assumed to liquidate unrelated trades merely because the trigger looks across the account. Check the measured balance or equity, the affected positions, and the pause behavior of each selected limit. EA-instance Hard Stop. Account-wide trigger.
The MT5 manual offers Classic broker-managed exits and Virtual EA-managed exits. Virtual mode needs the terminal and EA online to act; disconnecting removes that immediate protection. MT4’s listing emphasizes virtual exits. A VPS can help continuity, but it does not remove slippage, gaps or execution risk. Exit-type documentation.
Spread and scheduled-news filters can stop new entries under selected conditions. The public manual specifies the ForexFactory WebRequest URL for automatic news data. A filter being enabled is not proof that its feed is working, nor does an entry pause close every existing position. Verify the data permission and behavior on demo. News-filter requirements.
What performance evidence is available?
We did not establish a public, continuously tracked account that isolates automated SmartChoise results for either edition. In a June 5, 2026 reply to a critical MT5 review, the developer explained that his personal accounts combine the EA with manual entries and management, so presenting them as a pure EA signal would be misleading. The main listing directs readers toward accounts discussed in a Telegram group; we did not treat private-group material as independently verified evidence. Read the MT5 reviews and replies.
The developer’s documented NN Stress-Test Mode adds simulated adverse conditions inside the MT5 tester. It can help investigate sensitivity, but it remains a simulation; calling it a worst-case mode does not establish a bound on future losses. Forward observation should record build, settings, broker conditions, deposits, withdrawals and floating equity alongside closed profit. Developer stress-test guide.
MQL5 customer feedback: useful, but mixed
Positive MT5 feedback includes praise for developer responsiveness and configurable behavior. More critical feedback questions entry quality, extended drawdowns and the gap between backtests and observed trading. These are individual accounts of experience with different settings and versions, not a controlled performance comparison. The June criticism and the developer’s response are particularly relevant to the evidence gap described above. Read the MT5 reviews and replies.
Miko’s comment: It is ok (on xau always use %stop loss)
The stop-related disagreement elsewhere in the reviews deserves context: the developer lists configurable loss controls, while users can still choose recovery-led settings without a conventional tight stop on each position. The existence of a setting and the protection actually enabled are different things. Later MT5 releases also cannot retroactively validate an older review.
Third-party feedback remains separate from CheaperForex verified-purchase reviews. We have not imported these ratings into a WooCommerce score or treated the larger MT5 discussion as MT4 evidence.
Who is the better fit?
SmartChoise is more relevant to a trader prepared to study recovery exposure, keep the terminal reliable and test settings over changing market conditions. It is a weaker fit when the requirement is a simple fixed-loss trade process, minimal supervision or independently verified automated returns. Tight prop-firm limits also need specific testing; the presence of a Hard Stop is not a guarantee of compliance.
Choose between MT4 and MT5 primarily by platform and the controls you need, then evaluate that exact build. SmartChoise Battery and SmartBid are separate products from the same developer, not substitute evidence for these editions.
