Converters ask for your data
CSV exports contain customer lists, ledgers, mailing addresses. JSON contains API payloads with tokens. A ZIP is private by definition. Uploading any of them to convert a format is a bad trade.
File conversion that never leaves your browser
CSV to JSON, JSON to CSV, Markdown to HTML, and unzip — all processed locally on your own machine. Nothing is uploaded.
Drop files here or
Four conversions, no server
FileKit does four things and refuses to claim more. Each one parses the awkward cases properly rather than splitting on commas and hoping.
| Task | Engine | Notes |
|---|---|---|
| CSV to JSON | RFC 4180 parser | Handles quoted commas, BOM, CRLF and semicolon delimiters. |
| JSON to CSV | Flattened with dotted keys | Header is the union of all keys, so ragged rows stay aligned. |
| Markdown to HTML | CommonMark plus GFM, sanitised | Tables, task lists and fenced code. Script tags stripped. |
| Unzip | Individual or bulk | Password-protected entries are reported, never written corrupt. |
| Batch | 10 files free | The desktop version has no limit. |
| Network | Zero requests after load | There is no upload endpoint to send anything to. |
Why "no upload" is the whole point
The transfer costs more than the work
Parsing a 5 MB CSV takes a few milliseconds. Uploading it, queueing on a server, and downloading the result takes seconds and depends on a stranger staying online.
There is no server to breach
Static files on a CDN. No backend, no database, no upload endpoint. Not "we delete it after an hour" — there is nothing to delete.
Five things people get wrong with file conversion
Converting CSV with a naive splitter
Commas inside quoted fields break every row after the first bad one. Delimiters need a real RFC 4180 parser.
Publishing unsanitised Markdown output
Markdown permits inline HTML, so a converted stranger's file can carry script tags straight into your site.
Uploading archives to unzip them
A ZIP exists because its contents are bulky or private. Handing it to a website to get the same bytes back is indefensible.
Forgetting Excel's BOM
A Windows-saved CSV starts with a byte-order mark that silently becomes part of your first column name.
Assuming CSV keeps types
CSV is text. A number, a date and a string all become indistinguishable the moment they enter the file.
Questions
Are my files uploaded anywhere?
No. Every conversion runs inside your browser tab with a JavaScript library. After the page loads it makes zero network requests.
How many files can I convert at once?
10 per batch on the free web version. The desktop version has no limit.
Does it work offline?
Yes. Once the page has loaded you can disconnect the network and everything still works.
Is there a watermark or branding on my output?
No. Output files are clean, with nothing added.
What formats are supported?
CSV, JSON, Markdown, HTML and ZIP today. Nothing else is claimed.
Can it handle large files?
Into the tens of megabytes comfortably. The real limit is your browser's memory, not the tool.
Does it work with Excel CSVs?
Yes. Excel's UTF-8 BOM is stripped on input, and can be added back on output.
Is it really free?
Yes, permanently, for 10 files per batch. The $19 desktop version removes that limit and adds batch conversion.
Need it without a browser tab?
FileKit Desktop runs every conversion as a single offline file — no 10-file limit, batch conversion, and it never needs a network connection. One-time $19, no subscription.
See FileKit Desktop — $19