You realize the drill. Somebody drops a screenshot of a 3,000% backtest, slaps “AI-powered” on the title, and expects you handy over $500.
You have seen it 100 occasions. So have I. And I am uninterested in it too.
So here is what I’ll do as a substitute: a full Alpha Pulse AI efficiency breakdown. Each single quantity — the wins, the losses, the months that labored, and the one that did not. As a result of after testing dozens of “AI bots” on MQL5, I can inform you that an ideal fairness curve is virtually a confession that somebody is promoting you smoke.
If these numbers impress you, nice. If they do not, that is fantastic too. At the very least they’re actual.
Alpha Pulse AI Efficiency — Myfxbook Verified, April 7 2026
No cherry-picking. No “since inception” tips. No handy begin date. Simply the account because it stands proper now:
- Whole Acquire: +15.25%
- Month-to-month Common: +4.23%
- Max Drawdown: 8.60%
- Revenue Issue: 1.29
- Win Fee: 55%
- Whole Trades: 105
- Common Commerce Size: 2h 52m
- Beginning Steadiness: $7,046.58
- Present Steadiness: $8,123.22
Do not belief me. Confirm: Alpha Pulse AI Baseline on Myfxbook. Each commerce. Each dropping streak. Nothing hidden.
April has been significantly robust: +8.00% achieve, 80% win price on 5 trades. However I am not going to fake one good month defines the EA. I acquired too hyped about early outcomes as soon as earlier than — a few implausible weeks adopted by a few tough ones that introduced me again to actuality. That is how this works. The complete image is what issues.
55% Win Fee — And Why That is Really Good
In case you see “55% win price” and suppose that is mediocre, I get it. You have been conditioned by distributors promoting 90%+ win charges. These numbers come from two locations: cherry-picked backtests or martingale methods that may ultimately blow your account to items.
Here is the maths that truly issues:
- Common Win: $82.74 (2,212 pips)
- Common Loss: -$79.19 (2,228 pips)
- Expectancy per Commerce: +$10.25
The wins are barely bigger than the losses, and so they occur extra typically. PF 1.29 means for each greenback misplaced, $1.29 comes again. Throughout 105 trades, that is $1,076 in internet revenue on a $7K account.
Is it going to make you wealthy by Friday? No. Is it going to explode at 3 AM whilst you sleep? Additionally no. And proper now, in a market the place most EAs are designed to separate you out of your cash, actual Alpha Pulse AI efficiency like this — constant and verified — is the sting no one desires to promote you. As a result of it isn’t attractive sufficient for a touchdown web page.
Each commerce recorded. Each loss seen.
No screenshots. No curves with out context. Confirm the stay Myfxbook sign or verify the full product particulars.
8.60% Max Drawdown — The Quantity That Separates Survivors
Most merchants obsess over win price. Those who survive obsess over drawdown.
8.60% max drawdown on a stay account buying and selling Gold — one of the risky devices available in the market — means the chance administration is doing its job. Gold moved over $400 per ounce in March 2026 alone. The account stayed inside 8.6% of its peak.
Why does this matter greater than any return quantity? As a result of drawdown is the place self-discipline dies. A 30% drawdown on a stunning fairness curve means completely nothing if you cannot psychologically survive it. You may flip the EA off at -25%, miss the restoration, and declare “it did not work.” I’ve seen it occur a whole bunch of occasions.
8.6% is livable. If that drawdown already makes you uncomfortable, then you definitely’re outsized — and that is a threat administration drawback, not an EA drawback.
Finest and Worst — No Hiding
- Finest commerce: +$246.68 (Feb 9)
- Worst commerce: -$165.00 (Feb 16)
- Finest in pips: +8,178 (Mar 20)
- Worst in pips: -8,535 (Mar 23)
Sure, the worst commerce in pips is bigger than the very best. That occurs in actual buying and selling. If somebody reveals you a system the place each loss is smaller than each win, they’re both curve-fitting their backtest or outright mendacity. The system works as a result of the maths works throughout 105 trades, not as a result of each particular person commerce is ideal.
The Unhealthy Interval — As a result of There Was One
I will be sincere: early on, I had an excessive amount of hype. A few actually robust weeks made me suppose it was already working at full potential. Then got here a few tough weeks that introduced all the things again to earth.
That is not a failure. That is what actual Alpha Pulse AI efficiency seems like throughout time — not an ideal line going up, however a system that survives dangerous circumstances. If somebody sells you an EA the place each month is constructive, they’re both mendacity or working one thing that may ultimately implode. Good EAs survive dangerous markets. Nice EAs acknowledge them and cut back publicity.
Alpha Pulse AI does one thing that the majority EAs bodily cannot: earlier than each commerce, it sends actual market information to an AI mannequin — worth motion, volatility, session context — and will get again an precise evaluation. Not a shifting common crossover with “AI” within the title. An precise resolution from GPT-5.4, Gemini 3.1 Professional, or Claude Opus 4.6 about whether or not circumstances favor a commerce proper now.
When circumstances are rubbish, the EA typically stays out. That is why you see 2-3 trades per week, not 20. There are various fashions, many pairs, many timeframes… that is precisely why it’s a must to take it sluggish.
Longs vs Shorts — The AI Discovered One thing
- Longs Received: 50% (32/63 trades)
- Shorts Received: 61% (26/42 trades)
The EA has been notably higher at shorting Gold than shopping for it. This is sensible — XAUUSD has had sharp pullbacks through the broader rally, and the AI fashions appear to establish reversal circumstances extra precisely than continuation setups.
This is not one thing I programmed. It is one thing the AI realized from analyzing circumstances in actual time. A static EA would commerce each instructions with the identical logic. This one adapts — and the information reveals it.
One EA is a psychology take a look at, not a buying and selling plan.
Alpha Pulse handles Gold. However working a single technique alone means your complete fairness is dependent upon one instrument’s habits. Obtain the free USDJPY module and begin diversifying at zero value.
Who This Is For (And Who Ought to Keep Away)
This IS for you if:
- You perceive that 55% win price with constructive expectancy beats 90% win price with hidden martingale threat
- You need verified outcomes — not screenshots, not backtests, not “belief me bro”
- You possibly can let the EA run for months with out touching it after one dangerous week
- You need an EA that evolves as AI fashions enhance — with out handbook updates or new purchases
- You settle for that some weeks might be flat or adverse. That is not a bug. That is actuality.
That is NOT for you if:
- You want assured month-to-month returns. No one can assure that. If they are saying they’ll, they’re working a rip-off
- You may panic and switch the EA off through the first drawdown. At that time, you are not buying and selling — you are playing with further steps
- You count on 90%+ win charges. That is martingale territory, and you understand how that ends
- You need straightforward cash. This is not straightforward cash. There’s nonetheless threat. If somebody offered you the concept bots print cash, they scammed you
What’s Really Beneath the Hood
Actual AI buying and selling means passing actual market info to a mannequin and having that mannequin truly take part within the resolution. That is it. That is the definition. Every part else is advertising and marketing make-up.
Earlier than each commerce, Alpha Pulse AI sends present market information to considered one of these fashions:
- GPT-5.4 — OpenAI’s newest. Robust analytical depth
- Gemini 3.1 Professional — Google’s mannequin. Quick processing, free API tier (as much as $300 in credit)
- Claude Opus 4.6 — Anthropic’s deepest thinker. Finest for multi-layered evaluation
- Grok 4.20 — xAI’s mannequin. Sample recognition targeted
You select which mannequin to run. You turn everytime you need. When these suppliers launch higher fashions, your EA makes use of them mechanically. That is the auto-evolution benefit — since Gemini 3.1 Professional and GPT-5.4, the evaluation high quality has improved dramatically in comparison with earlier variations.
The presets have optimized the frequency of API calls and the way the evaluation is structured. This issues as a result of API prices are actual — and I would reasonably be sincere about prices than conceal them like everybody else.
Price Transparency
- EA Value: $397 one-time. No subscription. No recurring charges. No “premium tier” upsell
- AI API Price: $0/month with Gemini free tier, or $2-5/month with premium fashions at regular buying and selling frequency
- VPS: $15-30/month beneficial for constant execution (ForexVPS works effectively)
Whole ongoing value: $15-35/month. On a $7K account making 4% month-to-month, that is coated within the first commerce. On a smaller account, use Gemini free tier and skip VPS prices till you scale. There’s an actual value drawback with AI buying and selling — most distributors conceal it. I would reasonably you understand upfront.
FAQ
Can I confirm these outcomes independently?
Sure. Each quantity on this publish comes from the stay Myfxbook sign. Each commerce, each drawdown, each dropping streak. Nothing hidden. Proof with out context is nugatory — that is why I am exhibiting you context.
Why solely 105 trades? Is not {that a} small pattern?
It is sufficient to determine a sample, not sufficient to show lifetime consistency. I feel we’re on the level the place, if it isn’t but persistently worthwhile long-term, it’s extremely shut. At 200+ trades, the statistics grow to be far more sturdy. I will preserve publishing updates.
What occurs throughout excessive volatility occasions?
The AI fashions analyze volatility as a part of their resolution. Throughout March 2026 market turbulence, the EA lowered commerce frequency and saved drawdown managed. That is the distinction between a static EA and one that truly reads circumstances.
What dealer do you suggest?
IC Markets for uncooked spreads on Gold, or Axi Choose if you wish to scale capital with out problem charges. As an energetic Axi affiliate, I can escalate any account situation on to my supervisor — decision in days, not weeks.
Can I run this on a funded account?
Sure. The EA’s conservative threat settings (2% default) and managed drawdown (8.6% max noticed) make it appropriate for many funded account packages. Drop to 0.5-1% threat for stricter prop agency guidelines.
Sources
105 trades in. The route is constructive and the chance is managed. Observe alongside and see the place we’re at 200.
