Pick the format your scanner, retailer or system expects. Each guide covers what the format is for, what data it accepts, and how to print it so it scans.
Code 128
Code 128 encodes all standard keyboard characters and packs digits two per symbol, which makes it one of the most compact linear barcodes. It always contains a hidden check character that scanners verify automatically.
Used for: Shipping labels, warehouse and inventory systems, asset tags, and anywhere you control both the label and the scanner. The GS1-128 variant carries structured supply-chain data (that variant isn't generated here).
Data requirements: Letters (upper and lower case), digits, spaces and common punctuation. Long runs of digits produce a shorter barcode than letters do.
Common mistakes:
- Encoding very long text. Every character adds width; beyond about 20–30 characters handheld scanners struggle.
- Using it for retail products. Shops expect EAN-13 or UPC-A with a GS1-registered number, not Code 128.
Printing and scanning: A module width of 0.25–0.5 mm and a bar height of at least 15% of the barcode's length (and no less than about 6 mm) suit most handheld scanners. Keep at least 10 modules of blank space at each end.
More about Code 128, with the generator set up for it
Related formats: Code 39, Code 93, Data Matrix
Code 39
Code 39 is one of the oldest alphanumeric barcodes. Each character is made of nine elements, three of them wide. That makes it easy to print, but wide: it needs far more space per character than Code 128.
Used for: Industrial part marking, ID badges and legacy systems. The US Department of Defense (LOGMARS) and parts of the automotive industry have long used it.
Data requirements: Capital letters A–Z, digits 0–9, space, and - . $ / + %. Lowercase needs "Full ASCII" mode, which many scanners must be configured for, so ScanHatch doesn't generate it.
Common mistakes:
- Typing lowercase letters. Standard Code 39 has none; use Code 128 instead.
- Turning on the optional mod 43 check character without setting up the scanner. The scanner will then return an extra character at the end.
Printing and scanning: Because the format is wide, keep data short. Its tolerance for print variation makes it a good fit for dot-matrix and thermal-transfer printers on basic equipment.
More about Code 39, with the generator set up for it
Related formats: Code 93, Code 128, Codabar
Code 93
Code 93 was designed to replace Code 39 with a denser encoding. It supports the same basic character set but produces a noticeably shorter barcode, and it always includes two check characters.
Used for: Logistics and electronics labelling, and some postal applications. It is less widely deployed than Code 128, so check that your scanners have it enabled.
Data requirements: Capital letters, digits, space and - . $ / + %. The two check characters are calculated for you and are not part of the scanned value.
Common mistakes:
- Assuming every scanner reads it by default. Some ship with Code 93 disabled.
- Choosing it for new systems where Code 128 would be more widely supported.
Printing and scanning: Similar to Code 128: a module width around 0.33 mm, a quiet zone of 10 modules on each side, and a comfortable bar height.
More about Code 93, with the generator set up for it
Related formats: Code 39, Code 128
EAN-13
EAN-13 is the standard barcode on retail products outside North America. The 13 digits are a GTIN (Global Trade Item Number): a GS1 company prefix, an item reference and a final check digit.
Used for: Products sold in shops. Books use EAN-13 with ISBN numbers starting 978 or 979.
Data requirements: Exactly 13 digits including the check digit. Enter 12 and ScanHatch calculates the 13th.
Common mistakes:
- Making up a number. Retailers only accept GTINs licensed from GS1 (or a legitimate reseller of GS1 numbers). A barcode generator can't register a number for you.
- Printing it too small or truncating the bars. Retailers may reject labels below 80% of nominal size.
Printing and scanning: Nominal size is a 0.33 mm module (about 37.3 × 25.9 mm including quiet zones); GS1 allows 80% to 200%. ScanHatch's download is somewhat larger at the same module width because it adds a full quiet zone on both sides and prints the digits outside the bars. Keep the light margins clear, especially the 11-module zone on the left.
More about EAN-13, with the generator set up for it
Related formats: UPC-A, EAN-8, ITF-14
EAN-8
EAN-8 is a shortened retail barcode for packages too small for EAN-13. It holds a separate, scarcer type of GS1 number, not a shortened EAN-13.
Used for: Small retail items such as cosmetics, confectionery and pens, where an EAN-13 won't fit.
Data requirements: Exactly 8 digits including the check digit. Enter 7 to have the check digit calculated.
Common mistakes:
- Chopping digits off an existing EAN-13. EAN-8 numbers are issued separately by GS1.
- Using EAN-8 when an EAN-13 would fit. GS1 only issues EAN-8 numbers when space is genuinely limited.
Printing and scanning: Same module-width rules as EAN-13 (0.264–0.66 mm). Nominal size is roughly 26.7 × 21.6 mm including quiet zones; ScanHatch's download is somewhat larger because it adds a quiet zone on both sides and prints the digits outside the bars.
More about EAN-8, with the generator set up for it
Related formats: EAN-13, UPC-E
UPC-A
UPC-A is the 12-digit retail barcode used in the United States and Canada. It's identical in structure to an EAN-13 that starts with 0, which is why scanners often report it as a 13-digit number.
Used for: Retail products sold in North America. Scanners worldwide read it.
Data requirements: Exactly 12 digits including the check digit. Enter 11 to have the check digit calculated.
Common mistakes:
- Expecting a scanner to return exactly 12 digits. Many report the 13-digit GTIN form with a leading 0; both refer to the same product.
- Using an unregistered number for products you plan to sell through retailers.
Printing and scanning: Nominal module width 0.33 mm, allowed 80–200%. Keep 9 modules of clear space on each side and avoid shortening the bars.
More about UPC-A, with the generator set up for it
Related formats: UPC-E, EAN-13
UPC-E
UPC-E compresses a UPC-A number to 8 digits by removing zeros, producing a barcode about half the width. Only UPC-A numbers with zeros in specific positions can be compressed.
Used for: Small packages in North America, such as cans and single-serve items.
Data requirements: Number system 0 or 1, six digits, and a check digit that's calculated from the expanded UPC-A number (not from the 8 digits directly). You can also enter a full UPC-A to see whether it can be shortened.
Common mistakes:
- Calculating the check digit from the 8 UPC-E digits. It must be calculated from the expanded 12-digit UPC-A.
- Trying to compress a number without the right zeros. Not every UPC-A has a UPC-E form.
Printing and scanning: Same module-width guidance as UPC-A. Scanners usually return the expanded UPC-A (or 13-digit GTIN), not the 8 printed digits.
More about UPC-E, with the generator set up for it
Related formats: UPC-A, EAN-8
ITF (Interleaved 2 of 5)
Interleaved 2 of 5 encodes digits in pairs: one digit in the bars and the next in the spaces between them. That makes it compact for numbers and forgiving of rough printing.
Used for: Warehouse and distribution labels, cartons and industrial applications where only numbers are needed.
Data requirements: Digits only, and an even number of them. There's no built-in check, so a partial scan can be misread as a shorter valid number.
Common mistakes:
- Using very short codes. Scanners often refuse ITF shorter than 6 digits to avoid misreads; set a fixed length in the scanner if you can.
- Using it where letters might be needed later.
Printing and scanning: Bearer bars (a border above and below) help prevent partial reads on corrugated cardboard. The ITF-14 format adds them for you.
More about ITF (Interleaved 2 of 5), with the generator set up for it
Related formats: ITF-14, Code 128
ITF-14
ITF-14 is Interleaved 2 of 5 carrying a 14-digit GTIN for a group of products, usually printed with a thick bearer border so it survives printing on corrugated board.
Used for: Outer cartons and cases in the supply chain, rather than individual items sold at a till.
Data requirements: Exactly 14 digits including a GS1 check digit. The first digit is typically a packaging indicator (1–8), followed by the product's company prefix and item reference.
Common mistakes:
- Using the product's own EAN-13 number unchanged. A case of products usually needs its own GTIN-14 with a packaging indicator.
- Printing too small. ITF-14 is meant to be large, with a module width around 0.5–1 mm on cartons.
Printing and scanning: Print it large on cartons, with the full bearer border and a quiet zone of at least 10 modules. Scanning from a distance on a conveyor is normal.
More about ITF-14, with the generator set up for it
Related formats: ITF (Interleaved 2 of 5), EAN-13
Codabar
Codabar is a simple, self-checking numeric format from the 1970s. Every code starts and ends with one of the letters A, B, C or D, which some systems use to identify the type of item.
Used for: Library books, and historically blood banks, courier airbills and photo labs. Many of those have since moved to Code 128.
Data requirements: Digits 0–9 and - $ : / . +, framed by start and stop letters (A–D). Scanners usually return the start and stop letters too.
Common mistakes:
- Choosing start/stop letters your system doesn't expect. Check which ones your software uses.
- Choosing Codabar for a new system. Code 128 is more compact and more widely supported.
Printing and scanning: Tolerates low-resolution printing. Keep a quiet zone of about 10 modules and a comfortable bar height.
More about Codabar, with the generator set up for it
Related formats: Code 39, Code 128
MSI (Modified Plessey)
MSI (also called Modified Plessey) is a numeric-only barcode in which each digit is encoded as four binary bits. It usually carries one or two check digits calculated with mod 10 or mod 11.
Used for: Retail shelf-edge labels, warehouse bin locations and inventory systems, mostly in older installations.
Data requirements: Digits only. Choose the check-digit method your scanner is configured for: mod 10 is the most common.
Common mistakes:
- Choosing a check-digit type that doesn't match the scanner's setting, so every scan is rejected.
- Expecting phone apps to read it. Many, including ScanHatch's scanner, don't support MSI.
Printing and scanning: MSI is relatively wide per digit. Keep data short and use a dedicated barcode scanner for reading.
Related formats: Code 128, ITF (Interleaved 2 of 5)
Pharmacode
Pharmacode (Laetus) stores a single number as a pattern of thick and thin bars. It's a packaging-control code: machines on a packaging line check that the correct carton or leaflet is being used.
Used for: Pharmaceutical cartons, labels and patient leaflets on production lines. It doesn't identify products for retail.
Data requirements: A whole number from 3 to 131070. There is no check digit and no human-readable text as standard.
Common mistakes:
- Using it as a product identifier. For that, pharmaceuticals use GTINs in EAN-13 or GS1 Data Matrix.
- Printing the code upside down relative to the reader's expectation. It can read as a different number.
Printing and scanning: Print in the position and orientation your packaging-line reader expects. General-purpose phone apps usually can't read Pharmacode.
Related formats: Code 128, Data Matrix
Data Matrix
Data Matrix is a square 2D code that can hold dozens to hundreds of characters in a tiny space, with strong error correction so it still reads when partly damaged.
Used for: Marking electronic components and industrial parts, and on healthcare products. Those often use GS1 Data Matrix, a structured variant that isn't generated here.
Data requirements: Letters, digits, symbols and line breaks. More data makes a bigger symbol.
Common mistakes:
- Assuming a normal Data Matrix is a GS1 Data Matrix. Regulated healthcare uses need the GS1 format with specific data structure.
- Printing it too small for the camera or scanner that will read it.
Printing and scanning: Leave at least one module of blank space around the symbol. For phone scanning, 10–15 mm across or larger works well.
More about Data Matrix, with the generator set up for it
Related formats: Aztec, PDF417
PDF417
PDF417 is a stacked barcode: many short rows of linear codewords, which a scanner reads as one symbol. It holds several hundred characters with error correction.
Used for: Driving licences and ID cards, airline boarding passes and shipping documents.
Data requirements: Text including line breaks. Longer text adds rows and columns.
Common mistakes:
- Squeezing it too short. Rows need enough height for the scanner to track them.
- Cropping the start and stop patterns at the edges.
Printing and scanning: Give it a quiet zone of at least two modules and avoid very narrow modules; PDF417 needs a decent print resolution.
More about PDF417, with the generator set up for it
Related formats: Data Matrix, Aztec
Aztec
Aztec code is a square 2D code with a bullseye in the centre. Because scanners find it from the middle, it doesn't need a blank margin around it.
Used for: Rail and airline tickets and mobile boarding passes, especially where the code is shown on a phone screen.
Data requirements: Letters, digits, symbols and line breaks.
Common mistakes:
- Adding decoration over the central bullseye, which scanners use to locate the code.
- Choosing it where users will scan with generic apps. QR codes are more universally recognised.
Printing and scanning: No quiet zone is required, but a small margin doesn't hurt. Keep good contrast; screen brightness matters when shown on a phone.
More about Aztec, with the generator set up for it
Related formats: Data Matrix, PDF417