A spool of aerospace-grade ULTEM 9085 filament costs $250 to $400 per kilogram. The same ULTEM 9085 in pellet form — the raw feedstock injection molders buy by the ton — costs $40 to $70 per kilogram. The 6:1 to 8:1 price ratio between filament and pellet of the identical polymer is not a material difference — it is a processing margin: toll compounding the pellets into a precisely round filament with ±0.05 mm diameter tolerance, winding it onto a spool without tangling, drying it, vacuum-sealing it, and shipping it. A filament-based printer feeds this precision-engineered strand through a hot end and extrudes a 0.4-mm bead. A pellet printer skips the filament step entirely — it feeds raw pellets through a small screw extruder mounted on the print head and deposits a molten bead directly. The trade-off: pellet material costs one-sixth to one-eighth as much, but the extruder is heavier, more complex, and produces a coarser bead. This article compares the two feedstock approaches on material cost, throughput, part resolution, system complexity, and total cost of ownership — so you can calculate whether your production volume justifies the pellet path.
Why Filament Costs So Much More Than Pellets
Filament manufacturing is a secondary processing step. The polymer producer polymerizes and compounds the base resin — the same pellets injection molders use. A toll compounder or filament manufacturer re-extrudes those pellets through a precision die, pulling the strand through a water bath and a laser micrometer that controls diameter to ±0.05 mm or tighter. The strand is wound onto a spool at constant tension — a process that runs at 10 to 30 meters per minute and produces a spool with roughly 750 grams to 8 kilograms of usable filament. Any diameter deviation, ovality, contamination, or moisture absorption during winding creates a quality reject. The toll compounder charges for the material, the conversion labor, the quality control, the spool, the vacuum-sealed bag with desiccant, and the shipping — typically $150 to $300 per kilogram above the raw pellet cost.
Direct pellet printing eliminates every step between the railcar of resin and the 3D-printed part. The hopper on the print head holds raw pellets. A small single-screw or twin-screw extruder — essentially a miniature injection-molding barrel — melts, mixes, and pressurizes the polymer, then deposits a molten bead through a nozzle. The material cost per printed kilogram is the pellet price plus a few percent for the electricity to run the extruder heater.
Throughput: The Hidden Pellet Advantage
Filament-based FFF printers are throughput-limited by the hot end's melt capacity. A standard E3D V6-style hot end can melt 10 to 15 mm³/s of ABS — roughly 36 to 54 grams per hour. A high-flow hot end (E3D Volcano, Slice Engineering Mosquito Magnum) pushes 25 to 40 mm³/s — 90 to 144 g/h. A large pellet extruder on an industrial robot arm can deposit 1 to 10 kg/h — roughly 10 to 100 times the throughput of the fastest filament hot end — and the limiting factor becomes the motion system's ability to move the heavy extruder accurately, not the melt capacity. For large parts — tooling, thermoforming molds, boat hull plugs, architectural elements — the throughput difference can turn a 48-hour filament print into a 4-hour pellet print. For small, detailed parts, the heavier pellet extruder (5 to 25 kg versus 0.5 to 2 kg for a filament print head) limits acceleration and speed, and the throughput advantage disappears or reverses.
Part Resolution: Where Filament Wins
A pellet extruder deposits a 1.5 to 5.0 mm bead width, compared to 0.4 to 0.8 mm for a filament printer. Layer heights typically run 0.5 to 3.0 mm for pellets versus 0.1 to 0.3 mm for filament. The result is a coarser surface finish and a lower effective resolution — fine features below 3 mm, thin walls below 2 mm, and small holes below 3 mm diameter are not practical with pellet extrusion. Post-processing — machining the as-printed blank to final dimensions on a CNC router or mill — is standard for pellet-printed functional parts. This hybrid approach (3D print a near-net-shape blank, finish-machine the critical surfaces) is the dominant production workflow for large-format pellet printing and mirrors the casting-plus-machining workflow it often replaces.
Filament printing produces finer detail as-printed and can often forgo post-processing for non-cosmetic functional parts. For production applications where the as-printed surface is acceptable (brackets, fixtures, internal ducting), filament avoids the capital and labor cost of post-machining.
Total Cost of Ownership: The Crossover Calculation
The TCO equation has three terms: equipment cost, material cost per kilogram of printed part, and throughput (kilograms per year). A production filament printer costs $5,000 to $50,000. A pellet printer or robot-integrated pellet extrusion system costs $50,000 to $250,000. The pellet system costs 5 to 10 times more up front — but saves $200 to $330 per kilogram of printed material.
The annual volume at which pellet printing becomes cheaper is a simple break-even calculation. For ULTEM 9085 at $280/kg filament versus $55/kg pellet, saving $225/kg: a $150,000 pellet system that prints 500 kg/year saves $112,500 in material cost in year one — the equipment premium pays for itself in under 18 months. A $5,000 filament printer consuming 20 kg/year saves nothing by switching to pellet — the material savings never recover the equipment cost. The crossover is roughly 100 to 150 kg of printed output per year for high-cost engineering thermoplastics (PEEK, ULTEM, PEKK, PPSU). Below that volume, filament is the economically rational choice. Above it — or when printing large parts where filament throughput is physically too slow — pellet printing is the economic winner, provided the part geometry is compatible with the coarser resolution.
When should I stay with filament?
Filament is the right choice when annual throughput is below 100 kg, when the parts are small enough that the pellet extruder's mass and inertia degrade print speed and accuracy, when fine features or as-printed surface finish are required and post-machining is not viable, or when the production environment needs the simplicity of a spool-fed system — no hopper refilling, no feed-throat bridging, no screw-cleaning purges between material changes. Filament also provides a wider material selection: specialty filled and reinforced grades are available as filament from boutique compounders, while pellet printers need a minimum purchase quantity (typically 25 kg bags) that locks in the material choice more firmly than swapping a spool.
Pellet 3D printing is not a higher-quality process — it is a lower-cost process for large parts and high annual volumes. If you print under 100 kg of engineering thermoplastics per year, filament is simpler, more precise, and has a lower total cost. Above 100 kg — or when your parts are measured in meters — the pellet extruder's material savings and throughput advantage make it the economic winner. The decision is not about technology — it is about kilograms per year and meters of bead per minute.
