Drop any file. Any size.

safe · secure · no server · works offline · fast

…or paste a file you copied in Finder, or just start typing at the cursor.

Or try it now — txt, json, jsonl, js, html, md, mdx, yaml, xml, mermaid, csv, docx, css, svg, jpg, png, heic, pvt, mp3, wav, ogg, mp4, mpg, pdf, parquet, zip, gz, dmg, wasm, gguf, diff json, diff bin, camera or voice

↓ scroll for about & FAQ

Preview

Drop a CSV file, or use the file picker.

What is OmniViewer?

OmniViewer is a universal file viewer that runs entirely in your browser. Drop any file — JSON, CSV, logs, images, audio, video, or a mystery binary — and instantly see its raw content, plus a hex view with offsets and an ASCII panel. There is no upload and no server: the page reads only the bytes needed to paint your screen, so even multi‑gigabyte files open instantly and privately.

What you can do

View the raw content of any file, of any size

Edit any text file in place on the EDIT tab — a real editor for regular files, reflected live in every other tab; huge files edit line‑by‑line (double-click) through a tiny overlay, so even multi‑gigabyte files work, and download saves the result

Inspect any file in hex (offset · bytes · ASCII)

Paste a file you copied in Finder or Explorer, or type and paste text straight into the page — the format is recognized from the content (pasted JSON opens the JSON toolkit directly, and a dropdown tracks what you type)

Share a small file or snippet as a URL — the bytes travel in the link's #payload= fragment, never to a server (20 live examples)

Keep everything local: your file never leaves your computer

Format-specific tools — formatted JSON, CSV tables, media metadata and playback — are rolling out per format, powered by the same engine as fastjsonviewer.com and hugecsv.com.

Dedicated tools by file type

Every file gets the raw and hex views. These formats get a dedicated toolkit — open one directly:

JSON viewer & formatter — beautify, convert to YAML or CSV, and read structural stats

JSON Lines viewer — browse and validate every record of a .jsonl/.ndjson export, infer its schema, convert to CSV or a JSON array

CSV viewer — the whole file as a sortable, filterable table (21 GB / 100M rows tested), in-browser SQL, CSV‑to‑JSON, stats, export and split

HTML viewer & formatter — beautify, minify, and convert the DOM to JSON

Markdown viewer — render a live preview, tidy the source, export HTML, extract a table of contents and read document stats

MDX viewer — preview Markdown‑plus‑JSX, audit every component and import, and catch the ones that won’t resolve at build time

YAML viewer — convert to JSON, tidy the source, browse a structural tree and read document stats

XML viewer — pretty-print, convert to JSON, browse the element tree, run live XPath queries and read document stats

JavaScript viewer — format, compress, deobfuscate and analyze .js

MP3 viewer — inspect ID3 tags and MPEG frames, edit metadata, generate test files

MP4 viewer — play the video, read metadata, browse the box/atom tree, check stats and fix fast-start

PNG viewer — preview with zoom and a color palette, browse every chunk, and strip hidden metadata (EXIF/GPS, text, timestamps) in one click

See the full format & feature map for every format and the tabs it gets.

FAQ

Is my file uploaded anywhere?

No. OmniViewer is a static page with no server-side processing: your file is read directly by your browser and never leaves your computer.

How large a file can I open?

OmniViewer only ever reads the bytes it needs to paint the screen, so file size is effectively unlimited — it uses the same viewing engine as fastjsonviewer.com, which has been tested in-browser with a 40 GB file.

What file types are supported?

Every file gets a raw content view and a hex view today. Format-specific tools (formatted JSON, CSV tables, media metadata and playback) are rolling out per format, and unknown binaries always fall back to hex.

Can I type or paste instead of dropping a file?

Yes — the front page is an empty buffer. Just start typing at the cursor, or paste text, and the viewer (including the hex tab) treats it as content. The format is recognized from the content itself: pasting a JSON blob opens the /json toolkit directly and pasting prose opens /txt, while typing leaves the page alone and offers a “Continue in the … editor” button beside the tabs, tracking what the text looks like as you write. You can paste a file too: copy a PNG (or anything else) in Finder or Explorer and press Ctrl/Cmd+V — it opens exactly as if you had dropped it, and nothing is uploaded either way. Curious how? Read how we auto-detect the format of pasted text.

Can I share a file or a snippet as a URL?

Yes. For content up to 32 KB the toolbar shows a link action that mints a URL carrying the content itself in its #payload= fragment. The fragment never reaches a server or its logs, so the bytes stay private, and whoever opens the link gets the exact same bytes — text or binary — decoded locally. See omniviewer.org/share.html for 20 live examples.

How is this related to fastjsonviewer.com and hugecsv.com?

Same engine, same authors. OmniViewer is the universal front door for any file. hugecsv.com’s big-CSV workbench now runs here at /csv — the WebAssembly streaming parser, the full-file sortable and filterable table, the exporter and the splitter, tested to 21 GB and 100 million rows. fastjsonviewer.com remains the deep, format-specific tool for huge JSON.

Do you use analytics? Doesn't that break the privacy promise?

We run Google Analytics, but it only ever sees the page path — / vs /mp3 vs /json — never the URL's query or fragment, and never your file. There's no server to receive your bytes, and the viewer works fully offline. Here's exactly how, with the real code: how we use Google Analytics without breaking our promise.