Cosmos Orion EA Review (2026): A Gold H2 Three-Point Breakout EA at $299 With No Published Track Record

37.0 / 100

weighted verdict score on eight published criteria

0 signals

live signals for Cosmos Orion itself; the one linked on the Market page runs a different EA

$299

Market price on 29 Oct 2026, against 26.99 USD for our separate listing

Is Cosmos Orion legit? The verdict at a glance

Cosmos Orion is a real product with a real seller, a detailed description and a version history. We found no allegation of fraud or loss anywhere, and we do not make one. The question the score answers is narrower: what can a buyer check before paying? For this EA the honest answer is the description and little else. A three-point breakout on H2 is a coherent idea, and the risk controls it lists are sensible. But there is no tester report, no signal, no drawdown figure and no set of parameters on the public page. The scorecard and radar show how the total was built.

What Cosmos Orion actually is

Cosmos Orion is an Expert Advisor for MetaTrader 5, sold on the MQL5 Market by Chen Jia Qi of China (user name walter2008). The seller profile shows a rating of 4.1 from 285 votes, 5 products, 11 signals and a website at tradingvisionx.com. The product was published on 3 October 2026, showed version 2.1 and was last updated on 9 October 2026. It has 10 activations per licence, a free demo downloaded 415 times, 'Buy' as its only purchase option (there is no rental) and a rating of 5.0 from two reviews. The page also shows '16 purchases per month'. The tagline is 'Gold H2 Three-Point Structure Breakout EA' and it trades XAUUSD, or the broker's gold symbol with a prefix or suffix, on MT5 only.

The listing opens with a positioning statement: most gold EAs either guess the direction or hide risk behind grids and martingale, and Cosmos Orion does neither. It 'doesn't predict the market'. It waits for price to form a confirmed three-point structure, enters only on a real breakout and protects every order with a hard stop loss. Two phrases carry the design: 'No breakout, no trade' and 'Risk is calculated before every order'.

How we judged it, and who has something to sell

Evidence on EAs is never neutral. Here is every party near this product and what each stands to gain, ours included.

• The developer sells the EA at 299 USD, sells four other products (Gold Snap at 399 USD, Gold House at 749 USD and two TrendMaster FX versions at 638 USD each) and runs 11 signals, each priced at 999 USD a month with no subscribers. The purchase also carries an offer of a bonus EA that the developer uses internally. His other records are public, which counts for something, and they are also his sales pages.
• MQL5 and MetaQuotes host the Market, the reviews and the signals. The version history and review dates are the platform's own. They hold no performance data for this product because none exists.
• Buyers who wrote reviews are two, both giving 5 stars, both within three weeks of launch. One is a repeat buyer of the developer's Gold Snap. Testimony, not data, and neither can be checked from outside.
• ForexCopilotEA sells a listing for this EA at a far lower price. We gain if you buy it, which is why this verdict opens with that fact and applies the same rules as every other.

How the three-point structure breakout works

The listing describes the engine in plain terms, and it is worth restating because it is the only thing that can be judged. The EA reads closed H2 candles and looks for three confirmed pivots in sequence. A pivot is confirmed only on closed bars, so the EA does not repaint its structure. When a valid structure exists it places a Buy Stop or Sell Stop near the newest pivot, so a trade opens only if price actually breaks out. If the structure expires or is invalidated, the pending order is removed. No breakout means no trade.

This is a stop-entry breakout design, the same family as the products we cover under gold EAs. It buys strength and sells weakness after a range has formed. The standard weaknesses of that family apply. False breakouts, where price touches the entry, fills the order and reverses, are the main source of losses. Gaps and fast news moves can fill a stop order well beyond the planned price. And breakout systems tend to win less often than reversion systems and to rely on a few large winners. The vendor does not publish a win rate, but the reviews describe a 'very high' win rate on three trades, and a 5:1 reward on one trade. Those two descriptions do not agree with each other, which is a reason to wait for numbers.

The protections add more rules than the entry has. A profit lock pauses new entries until the next server day after two consecutive wins. A weekend guard removes pending orders before the Friday close and resumes on Monday. A macro window guard pauses around major US data releases. A volatility filter rejects markets that are too quiet or too wild. A maximum-spread input and a margin pre-check with a safety reserve stop orders that the account cannot afford. Taken together these explain the low trade frequency that both reviewers mention: they cut off many of the days on which a breakout might fire.

Cosmos Orion backtest and live results: what is published

The short answer is nothing. The Market page contains no backtest results, no win rate, no drawdown, no profit factor and no parameter set for Cosmos Orion. Our own store listing publishes none either. The text promises that the EA 'protects every order', and that it was 'built for real-world trading', but offers no figure to test those words against.

Read the reviews with the arithmetic in mind. Yulei Wang wrote on 23 October 2026 that after two weeks 'all three trades executed so far' made a profit. Robert L wrote on 9 October 2026, a week after buying on 3 October, that a 'first successful trade' had occurred. That is four trades in total across two reviewers, all winners as reported. A strategy with a true win rate of 40% would still produce four winners in four trades about 2.6% of the time (0.4 to the fourth power is 0.0256), and one with a true win rate of 60% would do so about 13% of the time. Four winning trades is therefore mild evidence at best that the win rate is high, and no evidence at all about how large the losers are. Both reviewers also say they are aware the EA trades rarely, which matches the design.

The linked signal 2388805 is not Cosmos Orion

The Market page links 'Gift EA live signal' at MQL5 signal 2388805, named 'HighestRisk'. A buyer could easily take it for Cosmos Orion's record, so we opened it and read it line by line. It is the developer's bonus EA, not the product on sale. We report it for two reasons: it is the only live number the page offers, and it shows how this seller's accounts behave. We do not use any of it to score Cosmos Orion.

Several things in that table separate it from the product on sale. Seven trades a week with an average hold of nine minutes is not the profile of a system that places pending orders on H2 structures and, per the reviews, trades 'not much'. The platform's notices say the account is newly opened and results may be random, that 80% of the growth came within two days (3.13% of 64 days), and that earlier checks in September and October flagged the same concentration. The monthly growth is shown as minus 2.36% against a total of plus 34.93%. The signal carries a 999 USD monthly price with no subscribers. The average loser is larger than the average winner, so its 66.66% win rate does not make a high profit factor. None of this is an accusation. It is a reminder that the one live number on the Cosmos Orion page is a different system's number, and that its quality is modest.

The Strict input: trading win rate for frequency

Version 2.1, released on 9 October 2026 and six days after launch, added one input. The version note reads: 'Added Strict Value: The higher the value, the higher the win rate, but the fewer trades. Minimum: 0, Maximum: 1.4.' This matters for how any future number should be read.

A selectivity input is a trade-off between hit rate and opportunity. Set low, the EA takes more structures, wins less often and trades more. Set high, it takes only the cleanest structures, wins more often and trades a few times a month. Any win rate quoted for Cosmos Orion is therefore a win rate at a particular Strict value. A buyer who runs 1.4 and compares the result with a track record produced at 0.6 is comparing two different strategies. The reviews do not say which setting they used, and the public page does not give a default.

There is a second consequence, a statistical one. The higher the setting, the fewer trades, and the longer a buyer must wait for a sample of any size. Suppose a setting produces about two trades a month, which is roughly the pace the two reviewers describe (three trades in two weeks and one in one week). A sample of 30 trades, which is still small, would take well over a year. This is the main reason the score is low: the design makes it slow to prove, and no proof has been published yet. It is also the reason a stricter filter should not be mistaken for a better product. It can make the equity curve feel smoother over a short window while leaving the question of the edge unanswered.

Cosmos Orion risk and drawdown

Cosmos Orion publishes no drawdown figure, which is the most important risk number for any EA. What we can assess is the design, and the design is mostly sound on paper. A hard stop loss on every pending order means each trade has a defined loss. The absence of grid, martingale and averaging down, if true, removes the failure modes that destroy accounts in the verified records we have seen. The margin pre-check with a safety reserve guards against an order the account cannot carry. The profit lock, the weekend guard and the macro window guard remove some of the largest gap and news risks. On those facts a Medium label is reasonable. It rests on stated controls, not on a record that shows them working, and that is why the risk score is only 4.5.

• Risk per order is chosen by the user, up to 5%. The five levels run from 1% to 5% of the account on each order. At L5, two losing trades in a row would cost about 10%, three about 14%, and a run of five about 23%, before any other effect. The seller's own risk warning tells new users to start with L1 or test on a demo account first. We agree.
• False breakouts. A stop-entry system pays for every breakout that fails. The hard stop caps each loss, but a cluster of failures in a range-bound month would still be a drawdown, and none is published to size it.
• Gaps and slippage. A stop order fills at the next available price. Around news, or at the Monday open, the fill can be worse than the planned level. The macro and weekend guards reduce this; they cannot remove it.
• Settings drift. A Strict value, a risk level and a basis for sizing (balance or initial balance) are all user choices. The EA a buyer runs can differ from the EA a reviewer ran.

The recovery calculator above shows why drawdown matters more than most buyers expect: a 20% loss needs a 25% gain to recover, and a 40% loss needs 67%. At 5% risk per order, a short losing streak lands in that territory. At 1%, it does not. The difference between L1 and L5 is therefore larger than the difference between most products on the Market, and the listing leaves the choice to you.

What buyers say about Cosmos Orion

The page shows a rating of 5.0 from two reviews and four comments. We read all of them. We weigh them as testimony, and the sample is too small to weigh them as more.

Yulei Wang, 23 October 2026, 5 stars: after two weeks of operation 'the win rate has been very high, despite a low trading frequency, with all three trades executed so far resulting in a profit'. He adds that 'drawdowns are inevitable for any strategy' and that he has 'greater confidence in the alignment between my backtesting results and actual live trading performance', and that the author has been responsive and helpful. Robert L, 9 October 2026, 5 stars: bought on 3 October, a week later 'already a first successful trade', accepts that the EA 'won't trade much' but expects a high-probability trade (5:1), cites a profit factor between 3 and 7 and a low relative equity drawdown as his two reasons to buy, and says he expects the backtest to align with live trading as it did with Gold Snap.

Three points follow. First, both reviewers refer to backtest results that the public page does not contain, which means they received something privately or ran their own tests. We have no way to check either. Second, both reviews describe a handful of trades, and neither describes a loss, so they say little about how the system behaves when it is wrong. Third, neither review has a developer reply on the page. The comments do show replies: on 4 October 2026 the seller thanked a purchaser, Sven Markus Weller, and said the manual and configuration files had been sent privately. A second buyer, DMP, asked on the same day for a set file and 'other needful information', and received an identical reply at 16:51. That is a courteous and fast reply, and also a standard one.

Cosmos Orion price: 299 USD on the Market, 26.99 USD here

The Market price on 29 October 2026 was 299 USD, the same on the English, Spanish, Russian, German, Japanese and Chinese versions of the page, with no rental option. The listing calls it an early-bird price for the first 10 copies and says buyers also receive one premium EA that the developer uses internally. The page said '4 remaining' when we read it, while it also showed 10 activations and '16 purchases per month'. Those numbers do not obviously reconcile, since activations are per licence and not per sale, so we do not draw a conclusion from them, but a buyer should not treat the countdown as exact. The same seller's other products cost 399 USD (Gold Snap), 749 USD (Gold House) and 638 USD (TrendMaster FX, in MT5 and MT4 versions), so 299 USD is the lowest price in his range, consistent with a new product.

Our price is about 9% of the Market price (26.99 divided by 299), a difference of 272.01 USD. A lower price is not a better product and is not a reason to skip the checks in this verdict; the unknowns are the same at either price. What the difference does change is the cost side of the decision: if the only goal is to test the logic on a demo or a small account, the lower price reduces the money at stake. The cost of the product is also small next to the other costs: a VPS, a low-spread ECN or Raw account and the 200 USD suggested deposit. The question of how our listing relates to the developer's own sales, including licence and provenance, is an open question for the store owner that we cannot answer here. Read the terms on our listing and ask us through the contact page if anything is unclear.

The free demo is the sensible first step at either price, with one caveat. According to the developer's own post, as quoted in search results, it runs only in the Strategy Tester. That lets you build your own backtest, and our guide to reading an MT5 backtest report covers what to check. It does not let you watch the EA place live orders on a demo chart. A backtest of an H2 pending-order system is meaningful only with good tick data and realistic spread, and our guide on backtest versus live results explains why.

Setting up Cosmos Orion: broker, account and VPS

The requirements are specific, and three of them can stop a purchase from working at all. The EA needs a hedging account; netting accounts are not supported. Many brokers offer both, so check the account type before depositing. It trades XAUUSD and accepts the broker's own symbol name with a prefix or suffix, and it should be attached to one chart only. And it computes its signals on H2 whatever chart it sits on, so the chart timeframe does not matter.

The listing recommends a low-spread ECN or Raw account and a 24/7 VPS, and suggests a minimum deposit of 200 USD. The first two follow from the design. A stop-entry system pays the spread at the moment of the breakout, when spreads are widest, and a maximum-spread input will skip trades if it is set tight. A VPS keeps pending orders under management through the week, which matters because the EA also removes pending orders before the Friday close and restores them on Monday. As for broker time, the EA detects the server's GMT offset and daylight saving from the daily gold trading break and falls back to a manual setting, by default GMT+2 with US daylight saving. That is convenient, and also a point to verify on the first week: confirm in the panel that the offset it displays is the one your broker uses, because the macro window guard and the weekend guard depend on it.

Will it last? The developer and the product's age

Cosmos Orion is 26 days old on the day we checked it, and it has had one update. That is too short a life to say anything about durability, so the sustainability score rests on the seller and on the design. On the seller: Chen Jia Qi has five products on the Market, a rating of 4.1 from 285 votes and signals going back at least 39 weeks. The product ratings are mixed in a familiar way: Gold Snap 4.52 from 27 reviews, Gold House 4.39 from 61, TrendMaster FX 4.22 from 79 and its MT4 version 4.44 from 16. A seller with a catalogue, a history of updates and a record of public accounts is a stronger bet than an anonymous one.

On the design: a three-point structure breakout on gold is a logic with an economic reason. Markets that form higher lows or lower highs and then break out are expressing a change in balance between buyers and sellers. But it is a regime-dependent edge. Breakouts on gold work in trending phases and suffer in ranges, and an EA that waits for confirmation will miss the first part of each move. The listing does not say how many years the logic was tested over or in which regimes, which is the information the sustainability criterion needs. Percentage-based exits, a profit lock and a volatility filter are signs of care. They are not evidence of a durable edge.

The score, criterion by criterion: 37.0 out of 100

The total is 37.0 out of 100. Evidence is scored 3.5, inside the 3 to 4.5 tier the method reserves for products with no verified record: checkable vendor material only. Without a grade A source it cannot rise above 4 in any case. Profitability is scored 2, in the 'negative or unmeasured' tier, because no figure for this EA is public; the buyers' numbers are claims, and we do not score claims as proof. Risk reaches 4.5 on the strength of the stated controls, which are specific and sensible, but the method does not give high marks to controls nobody has seen working. Fit, at 5.5, is the strongest criterion: the requirements are clear and workable, though several of them are hard conditions.

Who Cosmos Orion fits, and who it does not

• It fits a patient trader who wants a low-frequency, rule-based gold breakout system, has a hedging ECN or Raw account and a VPS, accepts that the record is empty, and will test on demo for a month or more at the lowest risk level before any live money.
• It does not fit anyone who needs published evidence before paying, who trades on a netting account or from a small balance where the minimum lot overrides the risk setting, or who wants steady weekly activity.
• It is a reasonable subject for study if you want to understand a pivot-structure breakout, since the panel draws the structures and entry levels on the chart.
• It is a poor choice for a buyer who is paying mainly because the early-bird price is about to end, or because a reviewer wrote that the first trades were winners.

How it compares with other EAs we have scored

Among the gold EAs we have scored, Cosmos Orion is the one with the clearest description and the thinnest record. Gold House, from the same seller, has 768 live trades behind its numbers; Gold Beaver and Pulse Engine each have a public signal to argue about. Cosmos Orion has a design to argue about. Its 37.0 reflects that gap, not a flaw in the idea. For how stop-entry breakout systems differ from grids, martingales and scalpers, see our strategy comparison. To see other options by type, browse gold EAs.

Questions to put to the developer before you pay

1. Can you publish a tester report for version 2.1 with the modelling quality, spread, Strict value and risk level used, and the number of trades and drawdown it shows?
2. Will you open a live signal for Cosmos Orion on its own account, separate from the bonus EA?
3. What Strict value and risk level did the reviewers' backtest figures use, and what is the default?
4. How many years and which market regimes was the three-point logic tested over?
5. What is the typical stop distance in dollars, so a small-account buyer can compare it with the 0.01 lot floor?
6. Will configuration files be published or included in the Market download, so that buyers do not need a private message?
7. How does the early-bird count relate to the 10 activations and 16 monthly purchases shown on the page?