Is theRegex free to use?

Yes — theRegex is completely free, with no sign-up, no paywalled features, and no artificial limits.

Does theRegex send my pattern or test string anywhere?

No. Everything runs entirely in your browser, including the actual regex matching, which runs in a background Web Worker — nothing you type is ever sent to a server. There is no server; this is a static page.

What happens if my pattern causes catastrophic backtracking?

theRegex runs every test in a background Web Worker with a 1.5-second timeout. A runaway pattern is stopped automatically instead of freezing your tab, and the Lint tab helps you spot the risk before it happens.

What languages does Code Export support?

JavaScript, Python, Java, PHP, Go and C#, with each language's flag equivalents and replacement syntax ($1 / $ vs \1 / (?P ) etc.) correctly translated.

Is there a size limit on the test string?

Matching is capped at 5,000 results per run as a safety net against pathological input, but there's no artificial limit on the test string's length beyond your browser's own memory.

Does the Recent Patterns list leave my browser?

No — it's stored only in this browser's localStorage, never transmitted anywhere, and you can clear it at any time from the History panel.

How reliable are the auto-correct and Generate features?

Both are best-effort heuristics, not guaranteed-correct solvers. Auto-correct only ever suggests a fix it has actually verified compiles. Generate always shows exactly which of your pass/fail examples the result satisfies, so you can judge it yourself before using it.

Free · Private · No Sign-up

The regex tester that explains, lints, and ships code with you.

g i m s u y

Did you mean ?

Test string

Check each line separately

Supports native replacement syntax: $1, $<name>, $&, $`, $'.

Replacement

Result

Static analysis of your pattern — catches common mistakes and performance footguns, with one-click fixes where it's safe to apply one automatically.

Ready-to-paste snippets for the pattern above, in six languages — flags and replacement syntax are translated per language.

Give a few strings that should match and (optionally) a few that shouldn't — this builds a best-effort pattern from them. It's a heuristic, not a guaranteed-general solution, so always check the results below and try it in the Matches tab.

Should match (one per line)

Should NOT match (optional, one per line)

Click any card to load it straight into the tester, with a matching sample test string.

Anchors

^Start of string / line

$End of string / line

\bWord boundary

\BNot a word boundary

Character classes

.Any character except line break

\d \DDigit / non-digit

\w \WWord / non-word character

\s \SWhitespace / non-whitespace

[abc]Any of a, b, or c

[^abc]Any character except a, b, c

[a-z]Any character in the range a–z

Quantifiers

*0 or more

+1 or more

?0 or 1 (optional)

{n}Exactly n times

{n,}n or more times

{n,m}Between n and m times

*? +?Lazy (as few as possible)

Groups & alternation

(...)Capturing group

(?:...)Non-capturing group

(?<name>...)Named capturing group

a|ba OR b

\1 \k<name>Backreference

Lookaround

(?=...)Positive lookahead

(?!...)Negative lookahead

(?<=...)Positive lookbehind

(?<!...)Negative lookbehind

Flags

gGlobal — find all matches

iCase-insensitive

mMultiline

sDot matches newline

uUnicode mode

ySticky

Both are best-effort heuristics, not guaranteed-correct solvers. Auto-correct only ever suggests a fix it has actually verified compiles — it never guesses blindly. Generate always shows you exactly which of your pass/fail examples the result satisfies, so you can judge it yourself before using it; for very varied examples it may fall back to an exact-match pattern rather than a general one.