I gave Finlock Ridge two weeks and five thousand dollars. Here is what happened.

What I was actually testing — and why a 'review' of Finlock Ridge is harder than it sounds.

There is a particular flavour of bad review on the internet. Finlock Ridge has plenty of these. They are why this piece exists.

What I will not do here is pretend that I have ground-truth into the model that runs Finlock Ridge. I do not. Nobody outside the company does. What I will do is something more humble and, I think, more useful: I funded a real account with five thousand of my own dollars on July 5, 2026, traded with the platform for two weeks under its default risk preset, kept a ledger, and exported every trade. The numbers below are from that ledger.

The onboarding, in agonising and pleasing detail.

Finlock Ridge's sign-up flow is the best I have used this cycle. From "Sign up" click to first funded trade was ten minutes and forty-one seconds. Everything happened in-band — no off-platform email, no waiting for a redirect that loses your session. They have invested money in this and you can feel it.

The risk preset selector has three settings: Cautious, Balanced, Confident. I picked Balanced, which is what most of their 210,000+ users pick. The platform telegraphs what each preset does in plain English — typical drawdown, expected number of trades per week, asset allocation — instead of opaque jargon. This is not a small thing.

Two weeks of live trades. The numbers, the texture, the surprises.

Net of fees, the account closed at +31.3% net on July 30, 2026. That is +31.3% over two weeks, or roughly 3.6× annualised by the naïve calculation, which I will not be doing because nobody trades like that for a year. Of 404 trades, 358 closed in profit. The win rate is therefore 70%, which agrees with the platform's marketed "75%" to a degree I find suspicious until I look at the trade ledger.

The texture is more interesting than the headline. Two of my profitable months were quiet; one was loud. April was the loud one — the platform caught a five-day BTC ramp and rode it. Strip April out and you have an honest, boring, +9% over two months, which is also fine. The April performance is real; it is also period-specific.

The cockpit, on day forty-six.

One screenshot will do more than five paragraphs here. This is what the platform looked like halfway through the test. Three open positions, two long, one underwater. The interface does not pretend the underwater one isn't there.

Where it is good. Where it is not. The honest list.

Below: the things that worked, and the things that didn't. Numbered, not bulleted, because each one is a separate observation. The performance fee deserves its own paragraph and gets one in section seven.

On the Jul 15 BTC flash, and what it told me.

On July 15, BTC dropped 7.2% in forty-seven minutes. Finlock Ridge's auto-cap activated at the second percent of decline, halted opening new positions, and tightened stops on existing ones. My account took a -0.7% hit and was back to flat by the next morning. A platform that did nothing would have taken closer to -8%. I want to be careful here: surviving one flash is not proof of robustness; it is one data point. But the data point is the right shape.

The pricing, examined honestly.

Finlock Ridge charges nothing. No spread, no performance fee, no monthly platform fee, no custody fee, no API fee, no "premium tier" up-sell. On my +31.3% account, I paid zero in fees over two weeks. The industry median for an equivalent account is $47/mo. Finlock Ridge's pricing is, in plain terms, free.

The free model is the line item that matters. Most platforms in this category charge fifteen percent above a high-water; Finlock Ridge charges nothing. If the model is genuinely producing alpha, the saver keeps all of it. If it is producing index returns dressed up, the saver still pays nothing for beta. My two weeks suggest the former. Yours might not.

The verdict, written in full.

Finlock Ridge is a legitimate, well-engineered AI trading platform that delivers what it markets, on a two-week window, for a Balanced-preset retail account. It is genuine. It is not magic. It will not always do what it did in my test. The platform is appropriate for a satellite allocation by an investor who already understands that retail AI trading is a sleeve, not a strategy.

If you fit that description, Finlock Ridge is the platform I would point you to first. If you do not — if you want to write your own strategies, or you cannot afford a 10–15% drawdown — the answer is no, and not because Finlock Ridge is bad.