ZPL vs. PDF Label Printing Speed: Zebra ZD621, Brother QL-1110NWB and A4 40-Up (2026)

A warehouse operator clicks Print, waits for a PDF viewer and print queue, and watches a fast thermal printer pause between labels. Sending the same SKU through a printer-language workflow can feel almost immediate. The difference is often how quickly usable print data reaches the printer—not how fast its motor turns.

Native barcode commands can reduce rendering and data-transfer work, particularly for repeated SKU labels. Two-across rolls also produce two labels per feed cycle. However, a fast laser printer printing 40 labels per A4 sheet can have higher continuous throughput. First-label response and large-batch output deserve separate comparisons.

This guide uses manufacturer specifications checked on September 23, 2026. The throughput figures below are calculations, not a physical printer benchmark. No manufacturer-sourced, controlled ZPL-versus-PDF timing comparison was found for these exact jobs.

What the manufacturers actually specify

Conversions use the example layouts below: two labels per roll row at a 28 mm pitch, or 40 occupied positions per A4 sheet. Compatible two-across Brother stock has not been verified; its 471-label figure is geometry only. These are theoretical rates, not measured PDF throughput.

Sources: Zebra ZD621 technical specifications, Brother QL-1110NWB specifications, HP OfficeJet Pro 9020e datasheet, and Brother laser specifications. The laser's 48-page figure is for US Letter, not A4.

These rates exclude important parts of a warehouse job: opening files, selecting media, waiting for a first page, replenishing stock and collecting output. They also use different measurement methods. A millimetres-per-second rating is not directly comparable with pages per minute until the label layout is specified.

Two-across thermal rolls: convert feed speed into labels per minute

Here, two-across means two labels side by side across the roll, not two labels one after another. Assume 45 × 25 mm labels, with 25 mm in the feed direction and an illustrative 3 mm gap between rows. The row pitch is therefore 28 mm. Measure your actual gap before applying this calculation.

Labels/minute = feed speed in mm/s × 60 × labels per row ÷ row pitch in mm

For a ZD621 at the published maximum:

• 203 DPI: 203 × 60 × 2 ÷ 28 = 870 labels/minute.
• 300 DPI: 152 × 60 × 2 ÷ 28 ≈ 651 labels/minute.

Ten thousand labels occupy 5,000 rows and 140 metres of media. Pure feed time is approximately 11 minutes at 203 DPI or 15 minutes at 300 DPI. These calculations use Zebra's rounded metric speeds.

Two columns double label output per millimetre compared with one column at the same row pitch, if the printer keeps the same feed speed. They do not double motor speed. More image data, heavier black coverage or insufficient buffering can prevent that ideal gain. Acceleration, roll changes, cutting, peeling and scan-quality requirements also increase completion time.

Selecting 300 DPI is a quality and layout decision: it does not automatically mean faster printing. For whole-dot barcode sizing, see our 203 DPI and 300 DPI label guide.

Why native ZPL can print SKU barcodes faster than PDF

Zebra's developer guidance recommends printer-language output, fewer graphics and reusable formats for performance. That supports investigating the software path when a printer hesitates despite a high rated speed. See Zebra's printing best practices.

Send the barcode's meaning, rather than its pixels

A typical PDF workflow goes through document creation, a viewer or print service, a driver, spooling and printer processing. Depending on the driver, barcode vectors may become raster images before transmission.

Native ZPL can instead describe a Code 128 barcode using ^BC, its position, module width and SKU data. The printer constructs the bars. Store a layout once, then send changing field values instead of repeatedly rebuilding the same artwork. Zebra documents this with its download-and-recall format example using ^DF and ^XF.

This removes avoidable work before printing. The largest improvement appears when rendering, transmission or repeated job setup was starving the printer of ready-to-print data. If both workflows already keep its buffer full, their sustained output can approach the same mechanical limit.

A raster payload explains the opportunity

Consider a monochrome 100 × 25 mm canvas covering one two-across row, excluding the feed gap. Using Zebra's native 8 or 12 dots/mm:

A few barcode and text instructions can describe simple SKU artwork much more compactly. However, these are raw image calculations, not actual PDF sizes or measured network traffic. PDF can contain compact vectors; drivers and printer protocols can compress graphics. Zebra's ^GF graphics command supports different transfer representations. A smaller payload does not translate into an equal proportional speed increase.

Three workflows that should not be conflated

1. Native ZPL barcode/text and stored formats: the printer receives instructions and variable data.
2. A bitmap inside ZPL, such as ^GFA: it bypasses a PDF workflow but still transfers image data.
3. PDF Direct: a compatible, configured Zebra printer receives and renders PDF itself, removing host-driver or conversion steps. See Zebra PDF Direct.

Consequently, “ZPL versus PDF” is incomplete without identifying the actual paths. Compare native commands, raster ZPL, driver-printed PDF and PDF Direct separately where available. A well-prepared PDF can print efficiently; a large image wrapped in ZPL can still be slow.

Brother QL-1110NWB: native printing without ZPL

The QL-1110NWB uses Brother Raster, ESC/P and P-touch Template, rather than Zebra ZPL. Brother also specifies DK die-cut and continuous-roll media. Its 101.6 mm printable width does not by itself establish compatibility with a generic two-across warehouse roll. See Brother's QL-1110NWB product specifications.

For repeated SKU labels, P-touch Template is the relevant comparison: prepare and transfer a template, then supply changing data. Its command reference includes Code 128 support. A Brother Raster job still carries image data; “native protocol” alone does not imply template-level efficiency. See the P-touch Template command reference.

Applying 110 mm/s to our hypothetical two-column, 28 mm pitch gives about 471 labels/minute, or 21 minutes of feed time per 10,000. This is geometry only, not an approved QL media configuration or an achievable-speed claim. We have not verified a matching two-across DK die-cut roll. Printing two designs on wide continuous DK stock also leaves them on one adhesive panel unless separately divided; a transverse cutter does not create a longitudinal die cut.

Brother's advertised 69 standard address labels/minute is another, specifically conditioned figure—text only at 300 × 300 DPI. It should not be doubled to predict this different job. See Brother's US product page and speed qualification.

A4 40-up: a laser can win the continuous-output calculation

With every position occupied, 10,000 labels require 250 A4 sheets:

Labels/minute = A4 pages/minute × 40

This table converts published rates; it does not measure adhesive-label performance. If you print PDFs on a label printer, actual printing speed can differ substantially from these best-case figures. It assumes continuous output, full sheets and no setup, replenishment or collection delay. The office models' ISO document rates are not promises for the required label-media setting. Use approved stock and its correct printing mode.

Handling can change the result substantially. HP specifies capacity for 50 label sheets in and 30 out, so a 250-sheet run needs attention. Brother directs label jobs through the MP tray and rear face-up output, and its specifications recommend removing label sheets immediately after exit. This is not an unattended 250-sheet plain-paper run. See the HP datasheet, Brother label-loading instructions and output handling notes.

For a continuous, fully populated batch, an office printer can therefore be competitive. For a succession of two-label orders, it still feeds a whole sheet each time, while the thermal workflow advances one short row. That difference matters more than the fully loaded headline rate at a packing station.

Why office printers can be less sensitive to PDF overhead

This is a workflow inference, not a rule that inkjet and laser printers ignore file complexity.

First, 40 labels share one A4 page's processing and paper handling. Our two-across example needs 5,000 row-sized pages for 10,000 labels, versus 250 sheet pages. Either can be submitted as one batch job; pages are not the same as operating-system jobs. Repeated per-page work has more opportunities to matter in the roll layout.

Second, office printers are built around document printing. The cited Brother laser lists PCL6, BR-Script3 and PDF 1.7 support. The HP inkjet lists PCL3 GUI, so it should not be described as a native PDF printer. Host rendering can still matter on either path. Their respective Brother and HP specifications identify those languages.

Finally, once page data arrives quickly enough, sheet transport, laser fusing or inkjet carriage passes and drying become the limiting work. Faster data preparation then yields little additional sustained output. Complex PDFs, oversized images, slow networks or weak drivers can still produce pauses. The useful question is whether data preparation or physical printing is the bottleneck.

Measure the job your warehouse actually does

Run two tests: short orders of two labels, and continuous batches of 100 or 1,000. Use identical SKU data and physical artwork. On the same Zebra, compare PDF and ZPL at the same DPI, darkness and configured speed; document any scaling or rasterization. For Brother, compare its actual driver and template workflows on verified DK stock.

Record:

• Click-to-first-usable-label and click-to-last-usable-label, timing physical output rather than the queue disappearing.
• Sent bytes, connection type, ready-versus-sleep state and cut frequency.
• Loading, output collection, peeling and reprint time, separating operator time from unattended time.
• Scan results from the beginning, middle and end; unreadable labels do not count as productive output.

Repeat runs and report a median with the media, driver and firmware versions. These measurements can substantiate a speed claim for your setup without inventing a universal “ZPL is five times faster” ratio.

FAQ

Is ZPL always faster than printing a PDF?

Native ZPL often removes work for simple repeated labels. Its benefit shrinks when PDF processing already keeps up with the print engine. Raster ZPL and PDF Direct also change the comparison.

Can the Brother QL-1110NWB print Zebra ZPL?

Its documented languages are Brother Raster, ESC/P and P-touch Template. Use a supported Brother workflow; selecting Zebra ZPL in software does not add printer compatibility.

Does two-across thermal printing always beat A4 40-up?

No. It improves output per feed distance and suits short batches. Fully occupied A4 sheets can give a fast laser higher throughput; label-mode speed and handling determine the actual result.

Set up the layout, then verify the whole workflow

Use BarcodeMate's Warehouse label workspace to set columns, margins, printer resolution and barcode module width, then inspect the layout before printing.

Start with a correctly sized, scannable label and time your real workflow. Then use our label-printing cost comparison to convert measured waiting, handling and waste into operating cost.