Barcode Generator

Continue in the tool

More pages for this tool

Code 128 internal labels

Code 128 for internal labels

Stockroom, asset and internal batch labels rarely follow retail's GS1 numbering: they use their own prefixes, hyphens and uppercase letters. Code 128 subset B accepts any printable ASCII character in that range, so an SKU like SKU-TOOLS-2026 becomes a scannable symbol without registering anything with an external numbering body.

How Code 128 builds the symbol, from START to STOP

The Code 128 standard defines three subsets, A, B and C. Subset A covers uppercase letters and control characters, subset C packs two digits into every symbol position, which shortens purely numeric codes, and subset B covers uppercase, lowercase, digits and punctuation across the full printable ASCII range from 0x20 to 0x7E, 95 possible characters. This generator fixes subset B, always opening with the START B code (value 104), because that is what typical internal labels need: hyphenated SKUs, batch codes with a slash and lowercase asset names, all without switching subsets mid-symbol.
The Code 128 check digit adds the START value (weight 1) to every data character's value multiplied by its position, 1, 2, 3 and so on, then reduces the result modulo 103. For SKU-TOOLS-2026 the subset B data values are 51, 43, 53, 13, 52, 47, 47, 44, 51, 13, 18, 16, 18 and 22; the weighted sum including START reaches 3196, and 3196 modulo 103 gives checksum 3. The finished symbol concatenates START B, the 14 data characters, checksum 3 and the STOP code (106), for 17 encoded positions total.
Because subset B only covers the 95 printable characters between 0x20 and 0x7E, it excludes accented letters: characters like ç, ã or é must be swapped for c, a and e before generating the label, and this generator caps input at 80 characters per symbol. The quiet zone, the bar-free margin on each end, stays fixed at 12 units regardless of the chosen scale, because that is the space a scanner needs to recognize where the START pattern begins before decoding a single character. Bar height accepts 48 to 180 units, enough for shelf labels and for larger thermal-printed asset tags.

Three internal labels, three symbol widths

Hyphenated alphanumeric SKU

Input: SKU-TOOLS-2026
Expected output: Checksum 3, símbolo com 17 posições (START B + 14 dados + checksum + STOP)

This is the generator's own default value, covering uppercase letters, digits and a hyphen inside a single subset B symbol.

Batch code with a space and a slash

Input: LOTE 08/26
Expected output: Checksum 76, símbolo com 13 posições

A space and a slash are both valid ASCII characters inside subset B, so neither needs an escape character.

Asset tag with a # character

Input: ATIVO#4471
Expected output: Checksum 68, símbolo com 13 posições

The # character sits inside the 0x20-0x7E range, so it works unmodified, unlike an accented letter.

Full tool FAQ

What is the difference between Code 128, EAN and UPC?

Code 128 is a dense and flexible symbology for internal operations. EAN and UPC are numeric retail standards and depend on valid commercial identification.

Is the check digit generated automatically?

Is SVG better than PNG for barcodes?

What is a quiet zone?

Can I invent an EAN for a real product?

Why does my printed barcode not scan?

Does the tool validate commercial registration?

Are barcode values sent to the server?

Frequently asked questions

Why use Code 128 instead of EAN-13 to label internal assets?

Because EAN-13 requires a 13-digit numeric structure tied to a company prefix registered with GS1, built for retail. Code 128 subset B accepts any letter, digit, hyphen or printable ASCII symbol, so you can bake location, category and sequence straight into the code text without registering anything externally.

Does the warehouse scanner need extra configuration to read Code 128?

On most modern laser or imaging scanners, no: they detect each symbology START/STOP pattern automatically and switch without manual setup. Older fixed-symbology scanners, configured to accept only one format, may need Code 128 turned on explicitly in the device menu.

Can I use accented letters like ç or ã in the label text?

No. Subset B only covers the 95 printable ASCII characters between 0x20 and 0x7E, and Latin accented letters fall outside that range. Swap them for c, a, e before generating, or the generator itself rejects the text with a format error.

Do longer label strings produce a noticeably wider barcode?

Yes. Every extra character adds one more weighted position to the checksum calculation and one more bar-space pair to the finished symbol, so a 20-character code visibly takes up more width than a 6-character one. For small shelf labels, it pays to keep the identification scheme as short as it can be.

Next steps

Continue in the tool

Open full tool

Code 128 internal labels

Code 128 for internal labels

How Code 128 builds the symbol, from START to STOP

The Code 128 standard defines three subsets, A, B and C. Subset A covers uppercase letters and control characters, subset C packs two digits into every symbol position, which shortens purely numeric codes, and subset B covers uppercase, lowercase, digits and punctuation across the full printable ASCII range from 0x20 to 0x7E, 95 possible characters. This generator fixes subset B, always opening with the START B code (value 104), because that is what typical internal labels need: hyphenated SKUs, batch codes with a slash and lowercase asset names, all without switching subsets mid-symbol.

The Code 128 check digit adds the START value (weight 1) to every data character's value multiplied by its position, 1, 2, 3 and so on, then reduces the result modulo 103. For SKU-TOOLS-2026 the subset B data values are 51, 43, 53, 13, 52, 47, 47, 44, 51, 13, 18, 16, 18 and 22; the weighted sum including START reaches 3196, and 3196 modulo 103 gives checksum 3. The finished symbol concatenates START B, the 14 data characters, checksum 3 and the STOP code (106), for 17 encoded positions total.

Because subset B only covers the 95 printable characters between 0x20 and 0x7E, it excludes accented letters: characters like ç, ã or é must be swapped for c, a and e before generating the label, and this generator caps input at 80 characters per symbol. The quiet zone, the bar-free margin on each end, stays fixed at 12 units regardless of the chosen scale, because that is the space a scanner needs to recognize where the START pattern begins before decoding a single character. Bar height accepts 48 to 180 units, enough for shelf labels and for larger thermal-printed asset tags.

Three internal labels, three symbol widths

Hyphenated alphanumeric SKU

Input: SKU-TOOLS-2026
Expected output: Checksum 3, símbolo com 17 posições (START B + 14 dados + checksum + STOP)

This is the generator's own default value, covering uppercase letters, digits and a hyphen inside a single subset B symbol.

Batch code with a space and a slash

Input: LOTE 08/26
Expected output: Checksum 76, símbolo com 13 posições

A space and a slash are both valid ASCII characters inside subset B, so neither needs an escape character.

Asset tag with a # character

Input: ATIVO#4471
Expected output: Checksum 68, símbolo com 13 posições

The # character sits inside the 0x20-0x7E range, so it works unmodified, unlike an accented letter.

Full tool FAQ

What is the difference between Code 128, EAN and UPC?

Code 128 is a dense and flexible symbology for internal operations. EAN and UPC are numeric retail standards and depend on valid commercial identification.

Is the check digit generated automatically?

Is SVG better than PNG for barcodes?

What is a quiet zone?

Can I invent an EAN for a real product?

Why does my printed barcode not scan?

Does the tool validate commercial registration?

Are barcode values sent to the server?

Frequently asked questions