How to estimate Glowforge print time with the machine off

The Glowforge app tells you the print time only after the machine is on, the file is uploaded and you have pressed Print. Owners have asked for an offline estimate since at least 2019. Until it exists, you can get within about 15% with a tape measure's worth of maths. This page shows the whole method, with real numbers, so you can do it by hand or let the free estimator do it in your browser.

1. What a Glowforge speed number actually means

In September 2017 Glowforge changed its speed settings from real units to a 100–1000 scale. The release notes came with a conversion spreadsheet, and owners in the Cutting speed units thread reduced it to two straight lines, one for cut/score (100–500) and one for engrave (100–1000). In the original units, speed 100 was 100 mm/min, cut 500 was 4,000 mm/min and engrave 1000 was 8,500 mm/min. Converted to inches per minute (in/min) and millimetres per second (mm/s), the lines are:

cut or score:  in/min = 4 + (speed − 100) × 153 / 400     (100 → 4,  500 → 157)
engrave:       in/min = 4 + (speed − 100) × 331 / 900     (100 → 4, 1000 → 335)
mm/s = in/min × 25.4 / 60

Two things follow. First, the scale has an offset, so doubling the number does not double the speed: cut 200 is about ten times faster than cut 100, not twice. Second, the published spreadsheet described the commanded head speed, not what the head achieves on a small shape, which is why you still need the overhead and calibration steps below.

2. Cut and score time: path length ÷ speed, plus overhead

For vector operations the laser follows every line in the layer once per pass. So you need the total path length of that layer. In Inkscape select the shapes and open Extensions → Visualize Path → Measure Path (set it to length); Illustrator users can use the Document Info panel or a path-length script. Then:

seconds = passes × (length ÷ speed) × (1 + overhead)

The overhead covers acceleration out of every corner and the travel between shapes. From comparing estimates against real prints, 15% is a good default for cuts with a handful of shapes; files with dozens of small holes or lots of text need 25–40%, because the head never reaches full speed on a 3 mm circle.

3. Engrave time: area ÷ (speed × line spacing), plus turnarounds

A raster engrave sweeps the head back and forth across the engraved region in lines. Glowforge sets the spacing as lines per inch (LPI); the time model wants the gap between lines in mm:

lines    = height of the engraved region ÷ spacing
seconds  = passes × ( area ÷ (speed × spacing) × (1 + overhead) + lines × 0.05 )

The head must slow, stop and reverse at the end of every line, which costs about 0.05 s per line regardless of width; that is why a tall, narrow engrave takes longer than a short, wide one of the same area. Use 10% overhead here. For "area", the honest figure lies between the filled area of the artwork and the area of its bounding box, because the head sweeps full lines across each shape's width even where there is nothing to burn. KerfPrice uses the average of the two; if you are doing this by hand and the artwork is a solid block, just use its area. Doubling LPI doubles engrave time, which is the single biggest lever on a long job.

4. Worked example: a 90 mm coaster

This is the sample file in the estimator (Inkscape, mm document): a 90 mm circle with four 3 mm holes to cut, a 60 × 60 mm square to score, and a 20 × 20 mm solid square to engrave. Settings: cut 170, score 500, engrave 1000 at 270 LPI.

If you load the sample in the estimator you will see 1 min 33 s, not 1 min 16 s: the estimator defaults every stroke layer to cut at speed 170, so the blue square is costed as a 21 s cut until you change its operation to score. That is deliberate; the tool never guesses a faster setting than the one you chose.

Add your own handling time (the estimator defaults to 3 minutes of setup per job and 20 seconds per part for loading and unloading), and you have the number the Glowforge app would have shown you, before the machine is on. For a batch of 20 coasters: 20 × 1:16 = 25 minutes of machine time, plus about 10 minutes of handling.

5. Calibrate once, then trust it

Every machine, material and file is slightly different, so the first estimate is a ±15% guess. Print the job once, note the actual time from the app, and divide: factor = actual ÷ estimated. Multiply every future estimate for that machine by the factor. In practice one calibration gets most owners within 5–10% on similar work, which is good enough to quote with a small buffer. The free estimator has a box for the actual time and applies the factor for you; the Pro plan remembers it per machine.

6. From minutes to a price

Time is the hard part; pricing is arithmetic you should not do in your head while a customer waits. The usual structure is: machine time × an hourly machine rate (owners quote $30–120/hour depending on market) + your labour (design, masking, weeding, packing) + material from sheet nesting + consumables, then markup, then gross up for Etsy or card fees so you net what you intended. For the coaster above at a $60/hour machine rate, $25/hour labour, 10 minutes of design, a $12 sheet of 20 × 12 in plywood charged by area used and 30% markup, the estimator (with the blue layer left at its default cut setting, 1 min 33 s) gives a list price of $12.57 on Etsy, of which $10.92 reaches you after fees. Change any input and the number updates; try it with your own file.

Sources and notes

• Speed formulas: Cutting speed units (Glowforge Owners Forum, Sept 2017), where palmercr posted the fit to Glowforge's 2017 conversion spreadsheet, and Is speed in mm/s? (2021) for later measurement attempts showing the real head speed on short moves falls below the commanded speed.
• The offline-estimate request: Can we please have the option to calculate cutting time whilst the GF is off? (2020) and Any options on an offline print time estimator?
• Glowforge Plus and Pro machines accept engrave speeds above 1000 after a settings change; this page uses the Basic scale. Treat every "typical use" setting as a starting point, never as a replacement for your own material tests.
• This page and the estimator are maintained by Lowtide Labs. The formulas are open: the engine that produces every number here is the same one that runs in your browser on the estimator page, and it never uploads your file.