Font Identifier
Identifying a font from a picture needs a fingerprint database. This page explains why, and gives you four methods that actually work.
Convert with this tool
or choose a file from your device — it is processed in this tab and never uploaded
Accepted: Any font file (for exact identification), or an image (for guidance only)
Files stay on your device. There is no upload, no queue and no watermark, so a conversion works even with the network switched off after the page has loaded.
Supported formats & processing
| Accepted input | Any font file (for exact identification), or an image (for guidance only) |
|---|---|
| Produced output | Exact identification when you have the file; guidance otherwise — no image analysis is performed |
| Processing | Web Audio API + canvas + native JavaScript encoders |
| Upload required | No — the file is read locally and never transmitted |
How to use this tool
- If you have the font file, use the metadata checkerThe name table gives the exact family, style, version and foundry. This is the only method that is certain.
- If you are looking at a web page, inspect the CSSRight-click the text, choose Inspect, and read font-family in the Computed panel. It names the font directly.
- If you have a PDF, list the embedded fontspdffonts file.pdf names every font embedded in the document.
- If you only have an image, use a dedicated visual matcherVisual font-matching services are built for exactly this and are far more reliable than any local guess.
- Confirm before you buyOnce you have a candidate name, compare the specific glyphs that vary most between similar typefaces.
About this tool
Identifying a typeface from an image is a shape-matching problem at scale. You need a database of fingerprints covering tens of thousands of fonts, a feature extractor that is robust to scaling, rotation, anti-aliasing and photographic distortion, and a ranking stage that returns candidates with confidence scores. That is a server-side service, not a browser tab.
A tool that compares a cropped photo against the few dozen fonts installed on your device would return confident-looking nonsense. That is worse than no tool at all, because it sends you down the wrong path.
What follows is the honest version: the four approaches that genuinely work, in order of reliability, starting with the one that gives an exact answer in seconds.
Features
- No fake confidence scores and no random guessing
- Exact identification via the font metadata checker
- CSS inspection method for identifying fonts on websites
- PDF font listing method with the exact command
- Points to visual matching services for image-only cases
- Explains the glyphs to compare when verifying a candidate
Browser limits on this page
The part that is genuinely impossible in a browser
Outlines. TrueType (TTF) and PostScript (OTF/CFF) are different glyph technologies, not two labels for the same data. Converting one into the other means re-drawing every glyph — a conversion driven by a rasteriser and curve fitter, not by a byte-level transformation. No browser API does that, and any page that offers it is either upsampling a picture or lying.
Naming. A browser can read a font's internal tables but cannot re-issue the digital signature and embedding metadata a font vendor expects, so a repackaged file is not a licensed re-release of the original typeface.
WOFF2. WOFF2 uses a pre-processing transform plus Brotli, and no browser exposes Brotli compression to script. Genuine WOFF2 output is not possible here without shipping a full Brotli+woff2 encoder in WebAssembly.
What this page does instead: it reads and validates the file you supply, states precisely which conversion is not supported and why, and points you at the desktop route that works (fonttools, fontforge, or the Google woff2 tools).
Font identification from an image
Matching a picture of text to a typeface needs a large fingerprint database and a shape-matching engine. Looking a crop up against the handful of fonts installed on one device would be guesswork, so no result is produced rather than a confident wrong answer.
With the font file in hand, the name table gives the exact family, version and foundry — use the Font Metadata Checker. From a web page, the computed font-family in developer tools is authoritative. From a PDF, pdffonts lists the embedded fonts.
Frequently asked questions
Can I identify a font from a screenshot at all?
Only with a service that maintains a fingerprint database of tens of thousands of typefaces. A browser cannot host that database or run the matching, so use a dedicated visual matcher rather than a tool that guesses.
How do I find the font a website is using?
Open developer tools, select the text element, and read the Computed styles panel. It lists the resolved font-family, and the Sources or Network panel shows the actual font files being loaded.
How do I identify fonts in a PDF?
Run pdffonts file.pdf on the command line. It lists every font referenced, including the subset prefix that tells you whether the font was embedded partially.
Which glyphs should I compare to confirm a match?
Look at the lowercase g, the uppercase Q tail, the numerals 1 and 3, the ampersand and the letter R leg. These vary most between typefaces that are otherwise similar.