No. The formatter is a lexer and layout engine written in JavaScript that runs inside this page, so your expression never leaves your browser and is never logged. You can verify it: open your browser devtools, switch to the Network tab, and format something — there are no requests. This matters more here than for most tools, because a DAX measure often encodes pricing, margin or compensation logic that should not be pasted into a third-party service.
What layout does it produce?
The one the Power Query Advanced Editor uses in its own examples: let on its own line, each step indented one level below it, and in with the final expression back at the outer level. Any step still too wide for the line width is broken at its outermost bracket with one argument per line, so a long Table.TransformColumnTypes becomes readable instead of merely being on its own line.
Does it handle step names like #"Removed Columns"?
Yes, and correctly — a quoted identifier is treated as one indivisible token, so the # never gets separated from the name. Since the Advanced Editor generates a #"..." name for almost every step you create in the UI, getting this wrong would break essentially every real query. Escaped quotes inside text ( "say ""hi""" ), record and list literals, ranges such as {1..10}, lambdas with =>, and intrinsics like #date and #table are all handled.
Is Power Query M case sensitive?
Yes, unlike DAX, and this is a common source of confusion. Table.FirstN is not the same as table.firstn, and each is a keyword while Each is not. Because of that, this formatter changes no capitalisation whatsoever — it only ever adjusts whitespace. Its self-check compares your input and its output exactly, without normalising case.
Could formatting change what my expression does?
It is designed so that it cannot, and it checks its own work. The tool never rewrites your logic — it splits the expression into tokens (names, literals, operators, comments) and only changes the whitespace between them. Before showing you anything it re-reads its own output and compares the token sequence against your input; if they differ in any way, it discards the result and shows an error instead. A refusal is rare, but it is deliberately preferred over output that looks fine and is subtly wrong.
Where do I paste the result back?
In Power BI Desktop or Excel, open Power Query, select your query, and choose Home then Advanced Editor. Replace the contents with the formatted version. Power Query does not preserve your layout — it re-serialises the query when saved — so treat formatting as something you do while reading or reviewing a query rather than as a permanent change to the file.