If you've spent any time looking at strategy builders, you've probably noticed the same pattern. There's a free version, a paid version, and a fairly large gap between them that the marketing pages are a bit vague about. StrategyQuant, EA Studio, the EAs you find on the MQL5 marketplace, Nebula. They all draw the line in slightly different places. So before you hand over money, it's worth understanding what you actually get for free and what you're really paying for.
Here's the short version. Free tiers almost always let you test. Paid tiers let you take the result somewhere. That's the line, and most of the time it's the right line.
What "free" usually means
A free tier normally gives you a working backtester and some way to generate or describe a strategy. You can load data, run an idea, look at the equity curve, and get a feel for whether something held up over the period you tested. That's genuinely useful. You can learn a lot about whether an idea has any legs without spending a penny.
Where free tiers tend to stop is at the door marked "now do something with it". You often can't export the code, can't compile it into an Expert Advisor, can't run it on a live or demo account, and can't combine several strategies into a portfolio. Some tools cap how many strategies you can generate. Others restrict you to one symbol, or hide the deeper analysis behind the paywall.
None of that is a scandal. Testing is cheap to give away. Codegen, compilation, deployment, and ongoing support cost the company real money, so that's where they charge. The honest question isn't "why isn't it all free". It's "is the free tier enough to answer my question before I pay".
What "paid" usually unlocks
Paid tiers are mostly about getting your work out of the sandbox. You export the source code so you own it. You compile it into a runnable EA. You deploy it to MetaTrader so it can actually trade a demo or live account, and you build portfolios so you're not betting everything on one strategy. Sometimes you get more generations, more symbols, deeper robustness analysis, and proper support when something breaks at 2am.
Different products bundle this differently. Some charge a one-off licence, some run a subscription, some sell you individual finished EAs. There's no universally "correct" model. What matters is whether the thing you're paying for is the thing you actually need. If you only ever want to test ideas and never deploy them, a paid tier might be money you don't need to spend yet.
The two things to actually watch for
Price and feature lists are the easy part. The two things that quietly cost people are harder to spot.
Opaque backtests. A pretty equity curve tells you almost nothing on its own. You want to know what spread and commission were assumed, whether the test accounted for slippage, what the data quality was like, and crucially whether the result was checked on data the strategy never saw during building. A strategy can look brilliant on the exact period it was tuned on and fall apart the moment conditions change. If a tool won't show you how it tested, or won't separate the building data from the checking data, be sceptical. That's not pessimism. It's just how this works.
Curve-fit marketplace EAs. The MQL5 marketplace and similar shops are full of finished EAs with gorgeous backtests attached. Some are honest. Plenty are tuned within an inch of their lives to look perfect on history, which is a very different thing from holding up going forward. You can't see the logic, you can't see how it was tested, and you can't tell luck from edge. Buying a black box because its past chart looks good is one of the most expensive habits in this space. At least when you build it yourself, you can see what you're trusting.
Where Nebula sits
For the sake of being upfront, here's how Nebula draws the line. The free download needs no account and no card. It opens on a single-pair backtester, so you can generate a strategy, test it, and look at how it did. Every strategy gets a robustness grade from 0 to 100. That grade is a measure of edge versus luck in the testing, looking at whether the result held up on data the strategy didn't see while it was being built. It is not a profit prediction and it never will be. A high grade means the strategy behaved consistently under stress in testing. It doesn't mean it'll make money.
The paid tiers (Starter, Pro, Ultimate) are where you unlock code generation, compiling to an MT5 Expert Advisor, deployment, and portfolios. Same line as everyone else, more or less. Testing is free, taking it live costs money.
On the prop-firm question, since people always ask: firms generally permit MT5 EAs, but the rules differ between them and you have to read yours. Challenge fees and the capital behind them are at risk. Passing is never guaranteed, and it depends on you and how you trade. No strategy builder, Nebula included, makes any strategy more likely to pass a challenge. Anyone telling you otherwise is selling.
So which do you need
If you're still learning what works, start free, anywhere. Test ideas. Get used to reading equity curves with a sceptical eye, and learn to spot a fit that's too good. Only pay when you've got something you actually want to run, and when you understand why you're paying for the part you're paying for.
If you want to try the testing-first approach without an account or a card, you can grab the free Nebula download and run a strategy through the grader yourself. Make your own mind up about whether it's any good.