Most lists of "free" backtesting software count anything with a free button. That is not the useful question. The useful question is whether the free tier lets you do the same thing a thousand times, because repetition is the entire mechanism by which backtesting teaches you anything. By that test, three of the six tools below are free and three are demonstrations.
Every limit below was read from the vendor's own page on 17 August 2026. Several differ sharply from how third-party reviews describe them.
| Tool | What the free tier actually permits | Verdict |
|---|---|---|
| FXAbsolute | All 15 instruments, 28.4M bars, unlimited sessions, no account | Genuinely free |
| MT5 Strategy Tester | Unlimited tests, real tick data, multi-core optimisation | Genuinely free |
| BacktestFX | 9+ years tick data with real spreads; account required, ads shown | Genuinely free |
| FX Replay Beginner | 2 sessions. 50 records each. 1 week retention. 1 indicator | Demonstration |
| TradingView Basic | No intraday replay at all — daily and above only | Demonstration |
| Forex Tester | Time-limited trial; no permanent free tier | Trial |
This is our tool, so weigh the recommendation accordingly and check the limits against the table above rather than taking the ranking on faith.
It replays 28,412,683 one-minute bars across 15 forex, index and crypto instruments spanning 2021 to 2026, bar by bar, in a browser. Drawing tools, limit and stop orders, an automatic trade journal and live statistics. No account, no download, no payment, and no cap on how many sessions you run.
What it does not do, stated up front because a free tool that hides its limits is worse than a paid one that does not:
Best for: learning price action, building screen time, rehearsing a discretionary method, prop-firm preparation.
Free with MetaTrader 5, and on pure capability it is remarkable for something that costs nothing: real broker tick data, multi-core parameter optimisation, genetic search, and built-in forward testing that reserves history the optimiser never sees.
The catch is categorical rather than a limitation to work around. It tests Expert Advisors. You cannot step through bars placing your own trades — visual mode shows you an EA trading, not a chart you can trade. If you do not write MQL5, there is nothing here for you.
One trap worth knowing before you trust a result: the modelling mode dropdown decides how honestly your strategy is simulated, and "Open prices only" discards all intrabar movement. Across 1,911,141 EURUSD one-minute bars, the average intrabar range varies from 2.83 pips at 14:00 UTC to 0.73 pips at 21:00 — a 3.9× spread. A mode that ignores that range is far more dangerous for a London-session strategy than an Asian-session one, and the reported "modelling quality" percentage will not warn you.
Best for: MQL5 developers, parameter optimisation, anyone testing rule-based systems.
The tool that beats us on an axis we cannot match, and pretending otherwise would make this whole page an advert. BacktestFX offers 9+ years of tick-level data with real market spreads, free, with TradingView charts.
Deeper history and spread modelling are exactly the two things our own free tier lacks. The costs are an account requirement and advertising, which is a perfectly reasonable trade for what you get.
Best for: anyone whose strategy depends on spread and who will not pay. If you are scalping and choosing between the free options, choose this one.
Two sessions, fifty records per session, one week of data retention, one indicator. The software above that tier is genuinely good — the workflow is the most polished of the browser tools — but the free plan cannot support learning, because learning is repetition and repetition is what is capped. Paid tiers are $17.99/month or $180/year (Intermediate) and $35/month or $350/year (Pro).
The free plan has no intraday bar replay. Daily timeframes and above only. For a forex trader wanting to rehearse intraday decisions, the free tier does not offer the feature at all. Paid depth: Essential gives roughly six months of one-minute history, Plus about a year, Premium all stored data, Ultimate adds seven days of tick replay.
A trial, not a free tier. Worth taking if you are close to buying — and the software is genuinely strong — but it is not an answer to "what can I use for free indefinitely".
| If you… | Use | Because |
|---|---|---|
| Trade discretionary price action | FXAbsolute | The others cannot do manual replay |
| Write strategies in MQL5 | MT5 Strategy Tester | Optimisation nothing free matches |
| Scalp under 10 pips | BacktestFX | Real spreads change the answer |
| Are on a Mac or a phone | FXAbsolute | MT5 is a Windows desktop install |
| Want to start in the next 60 seconds | FXAbsolute | No account, no download |
| Want the deepest free history | BacktestFX | 9+ years vs our 5 |
There is no reason to pick only one. All three are free, and they fail in different directions — manual versus automated, depth versus immediacy, spread modelling versus no account. Traders who are serious about this generally end up using two of them for different jobs, and the total cost is still nothing.
For manual discretionary practice, FXAbsolute — free with no account, 28,412,683 one-minute bars across 15 instruments from 2021 to 2026, and no cap on sessions. For coded strategies, the MT5 Strategy Tester, free with MetaTrader 5 and supporting real broker tick data. For free tick data with real spreads, BacktestFX, which requires an account and shows ads.
Yes. FXAbsolute imposes no session cap, no retention limit and no account requirement. The MT5 Strategy Tester is unlimited but tests only coded Expert Advisors. By contrast, FX Replay's free tier allows just 2 sessions with 50 records each and one week of data retention, and TradingView's free plan has no intraday replay at all.
Only on daily timeframes and above. As of 17 August 2026 the free Basic plan has no intraday bar replay whatsoever, so it cannot be used to rehearse intraday forex decisions. Intraday replay depth begins on paid tiers, with Essential providing roughly six months of one-minute history.
Yes. FXAbsolute runs entirely in a browser with no install and no account, so it works on Windows, macOS, Linux, iOS and Android. BacktestFX and FX Replay are also browser-based, though both require an account and FX Replay caps its free tier heavily.
Usually one of three: a capped free tier that permits only a few sessions, missing spread modelling because the data is mid-price OHLC, or no automation so every trade is placed by hand. Free tools are well suited to learning price action through repetition and poorly suited to validating strategies that depend on precise execution costs.
No. What learning requires is unlimited repetition on realistic data, and several free tools provide exactly that. Paying becomes worthwhile when your method depends on tick-level accuracy or spread modelling, or when you need automated parameter sweeps — not before the habit exists.
FXAbsolute gives unlimited repetitions on five years of real one-minute data across 15 instruments — the repetition that actually teaches you something.
Open FXAbsolute →