How much filament is left on a spool? 4 ways to tell
Which method to use
If you have a scale and know what the empty spool weighs, weigh it: it is the most accurate method and works for any brand. With no scale, or with the spool stuck in a holder, measure the diameter of the wound filament; it gives a reasonable estimate, and a much better one if you measured the spool when it was new. The AMS remaining bar only shows up for Bambu Lab filament with an RFID tag, and it is an estimate in percent. Logging what each print uses works for people who record everything, as long as failed prints get recorded too.
| Method | What you need | Accuracy | Limitation |
|---|---|---|---|
| Weigh and subtract the spool | a scale with 1 g resolution and the empty spool weight | high, with an empty spool of the same brand | the spool has to come off the holder |
| Measure the diameter | a ruler or calipers | about 7% either way without calibration; better if you measured it full | uneven winding throws it off |
| AMS estimate | Bambu Lab filament with an RFID tag | estimated percentage | does not work with other brands |
| Log what you use | a record of every print | good, if failures are included | errors add up over time |
Method 1: weigh it and subtract the spool
- Zero the scale with nothing on it, and set it to grams.
- Weigh the spool on its own, off the holder or the AMS.
- Subtract the weight of an empty spool of the same brand and the same spool model.
A math example: the scale reads 412 g and an empty spool of that brand weighed 162 g. That leaves 412 − 162 = 250 g. In 1.75 mm PLA, at about 2.98 g per meter, that is roughly 84 meters.
The weak spot is the spool weight. It varies a lot from brand to brand, and the same brand sometimes changes spools. Some published values:
| Brand (1 kg spool) | Empty spool | Source |
|---|---|---|
| Polymaker (cardboard) | 140 g ± 7 g | Polymaker's website |
| Overture | 155 g | PrusaSlicer profiles |
| Elegoo | 160 g | PrusaSlicer profiles |
| Sunlu | 162 g | PrusaSlicer profiles |
| Verbatim | 235 g | PrusaSlicer profiles |
| Hatchbox | 245 g | PrusaSlicer profiles |
| eSun | 265 g | PrusaSlicer profiles |
| Fiberlogy | 330 g | PrusaSlicer profiles |
Sources: Polymaker product page and the spool weight field in PrusaSlicer's filament profiles, maintained by Prusa Research (profile version 2.5.13, checked November 2026). Bambu Lab's wiki and store do not list an empty spool weight, and for brands not in the table, weighing is the way to find out. Between the brands in the table, the difference reaches 190 g, so using an "average" spool weight can throw the math off by more than a whole part. When you finish a spool, weigh it empty and write the weight on it with a marker.
Method 2: measure the diameter of the wound filament
The filament fills a ring between the hub and the outer edge of the winding. The volume of that ring is π ÷ 4 × (D² − d²) × L. But round filament does not fill the whole ring: there are gaps between the turns. Geometry puts the filled share between 0.79 (turns stacked in a square pattern) and 0.91 (turns nested into each other), and that range is the uncertainty of this method.
A math example, with a 100 mm hub, filament wound out to 140 mm, 60 mm between the flanges and PLA at 1.24 g/cm³:
- Ring volume: 0.7854 × (140² − 100²) × 60 = 452,389 mm³, or 452 cm³.
- Filament: 452 cm³ × 0.79 to 0.91 × 1.24 g/cm³ = 443 g to 510 g.
You can remove that uncertainty if you measured the diameter when the spool was new. The width and the filled share cancel out, and the math becomes left = net weight × (D² − d²) ÷ (Dfull² − d²). If this 1 kg spool measured 175 mm when full: 1,000 g × (140² − 100²) ÷ (175² − 100²) = 1,000 × 9,600 ÷ 20,625 = 465 g.
Method 3: the AMS estimate
On Bambu Lab's AMS, the estimate relies on the spool's RFID tag: the machine measures how much filament is fed out per turn of the spool and works out the remaining percentage, taking a complete 1 kg spool as 100%. The result shows as a small bar under the filament name. The feature has to be turned on in the AMS options, and the tag is then read twice, as explained on the AMS HT page of the Bambu Lab wiki (checked November 2026).
There are three limits. It only works with tagged Bambu Lab filament, so other brands get no estimate. The result is a calculated percentage, not a measured weight. And it starts from a complete 1 kg spool. On other changers that read the tag on the brand's own filament, check the manual for how the estimate works before trusting it with a long print.
Method 4: log what you use
Start from the new spool's net weight and subtract the Total weight the slicer shows for each print. On a multicolor print, use the Total for each color, which already includes purge and prime tower; the guide to multicolor waste cost explains those columns. Failed prints use filament too: subtract the part that actually printed. In MakerPrice 3D, the Stock tab does this math: you register each spool, and when you confirm a print the app subtracts the grams from the spool you used. Every so often, check against the scale.
Will the next print fit on what is left?
Compare the filament left with the slicer's Total for that spool, and keep a margin for what the math cannot see:
- The spool weight. If you weighed an empty one of the same brand, the error is small; Polymaker, for example, states a variation of 7 g either way. With a generic weight, the error can go past 100 g.
- The scale. A kitchen scale with 1 g resolution adds very little error.
- The filament left in the path. When the spool runs out, the piece inside the tube up to the extruder never gets printed. Each meter of 1.75 mm PLA is about 3 g; measure the path on your machine.
A math example: 250 g are left, weighed against an empty spool of the same brand, and the slicer needs 230 g. The 20 g margin covers the spool variation, the scale and about 2 meters of tube. With a generic spool weight, the same math gives no real safety.
PrusaSlicer has this comparison built in: fill in Spool weight in the filament settings, and the slice summary also shows the weight of the filament plus the spool, which you can compare straight against the scale. That field is where most of the values in the table above come from.
If it does not fit, you have options. On the AMS, AMS Auto-Refill switches to another spool when the first one runs out, as long as both are the same brand, type and color. On printers with a filament runout sensor, the print pauses so you can swap the spool, which may leave a mark on that layer. And the end of a spool works well for small parts such as keychains and test prints. To know what the leftover is worth, multiply the grams by the spool's cost per gram, as in the guide to cost per gram.
Frequently asked questions
How much does an empty Bambu Lab spool weigh?
Bambu Lab's wiki and store do not list it (checked November 2026). Since the brand's refills go on a reusable spool, weigh that spool once, empty, and write the weight on it. The same goes for brands that do not publish their spool weight.
How many meters are in 1 kg of filament?
In 1.75 mm, about 335 m of PLA, 327 m of PETG and 400 m of ABS. The math is the weight divided by the grams per meter, and grams per meter are 2.405 cm³ per meter times the density: 1.24 g/cm³ for PLA, 1.27 for PETG and 1.04 for ABS.
Is a kitchen scale good enough?
Yes, if it reads in 1 g steps and its capacity is above the weight of a full spool, empty reel included, which is over 1.1 kg. Zero the empty scale and weigh the spool on its own, without the holder.
My new spool weighed less than 1 kg plus the spool. Is filament missing?
Before deciding, weigh an empty spool of that brand: the spool weight you assumed may be wrong. If the gap is still there after that, on more than one spool, it is worth contacting the seller with photos of the scale.
Is the end of a spool that sat open for months still usable?
It depends on the material and how it was stored. If the filament pops at the nozzle, strings or leaves a rough surface, it has absorbed moisture and needs drying before use; the guide to drying wet filament covers temperatures and times.
