How to Print High Temperature PEEK and PEI: Proven Guide (2026)

How to print high temperature PEEK and PEI comes down to three numbers and one habit: a nozzle that holds 400°C or above, a build plate that stays above the glass transition temperature of the polymer, an actively heated chamber that keeps every layer warm, and filament that is bone dry when it enters the extruder. PEEK is semi-crystalline and melts near 343°C, so it needs the most heat and a slow, controlled approach; PEI is amorphous and prints at a slightly lower burden. They are not interchangeable, and treating them as such is the reason most first high-temperature prints fail.

I have watched a lot of people approach this in the wrong order — buying filament first, then worrying about the machine. Work the other way around. Confirm your hardware can hold temperature, get the supplier’s validated profile for your exact grade, dry the filament, and run a small coupon before you commit to a full part.

Table of Contents

What You Need

The short answer is that you need an all-metal hotend rated to 500°C, an enclosed frame with an actively heated chamber, a heated bed that reaches at least 120°C, a rigid motion system, and fume extraction. Without those five things, PEEK and PEI printing is not a settings problem you can solve in the slicer.

PEEK and PEI are two different jobs

PEEK (polyetheretherketone) is a semi-crystalline PAEK-family thermoplastic. Its properties depend on how much crystalline structure develops during printing, which means the thermal history of the part matters as much as the extrusion temperature. It holds continuous service temperatures around 260°C and shrugs off steam and harsh chemicals.

PEI (polyetherimide, sold by SABIC as ULTEM) is amorphous. It has no crystallisation phase to drive, so it prints more predictably and reaches high strength without a post-print anneal. The common grades are ULTEM 1010, a lower-flow version, and ULTEM 9085, the higher-flow grade most people start with.

As a rule, start with PEI 9085 and step up to PEEK once your machine holds temperature steadily through a long print. Several people who posted their conversion notes on the Ultimaker community forum made exactly that move, and their advice was consistent: fix the hardware before chasing filament brands.

The hardware checklist

Work down this list and confirm every line before you order filament.

  • All-metal hotend rated to 500°C with no PTFE in the melt zone. PTFE softens well below the temperatures PEEK needs, and liner degradation is the top hardware reason consumer conversions fail.
  • Actively heated chamber with a real heater and insulation, not a warm box. Around 80°C is the practical ceiling for most converted machines because motors and electronics start complaining above it.
  • Heated bed reaching 120°C or more, verified with an independent thermometer rather than the machine’s own display.
  • Heater cartridge with enough wattage to reach and hold nozzle temperature. The 25W cartridge common on consumer machines plateaus around 400°C and triggers firmware timeouts; a 40W 24V cartridge fixed that for one converter.
  • Thermocouple or sensor rated for the range. PT100 sensors were reported failing repeatedly at 350–400°C in that same conversion, which meant swapping to a K-type thermocouple with a suitable amplifier.
  • Insulated heat block. Ceramic tape wrapped over the block and secured with polyimide tape cut heat-up time noticeably at 400°C and above.
  • Filament drying system with a way to dry in the spool or a dry box you can feed from.
  • Fume extraction running before the heater comes on.

What each chamber temperature tier unlocks

This is the fastest way to understand what a machine can actually do.

Chamber temperatureWhat it unlocksTypical materials
No active chamber heatingConsistent geometry at small part sizes only; layer bonding is weak on anything tallStandard thermoplastics
Around 80°CReliable small PEI and PEKK parts; PEEK at reduced size and slower speedsPEI 9085, PEKK, PPSU
Around 100°CLarger PEI parts, better crystallisation on PEEK, fewer warping failuresPEI 1010, PEEK
Around 150°CPEEK approaching in-situ crystallisation; the glass transition region of the polymer sits herePEEK
Around 200°CFull crystallisation during the print and the largest reliable PEEK partsPEEK, filled PEEK grades

One more caution before you buy. Vendor datasheets describe resin, and filament specs describe extruded strand. Printed parts are a third thing entirely. Follow the datasheet your filament supplier publishes for FFF, and where possible follow a profile validated on your machine class.

Step-by-Step: How to Print High Temperature PEEK and PEI Safely

The workflow below runs in the order that saves you time. Skipping the early steps is why people end up with clogged nozzles and lifted first layers after hours of printing.

1. Identify the polymer and machine limits

Read the spool, not the listing. VICTREX AM 450 FIL is PEEK. ULTEM AM1010F is PEI. Some Ultem-class filaments are copolyimides with different flow and thermal behaviour, and one Reddit thread about high-temperature food-contact parts turned on exactly that confusion.

Then check your machine’s validated range against the supplier’s requirements. For VICTREX AM 450 FIL the published conditions call for a melt window of 340–450°C, a build space of at least 150°C, and a bed above 150°C. For ULTEM AM1010F the figures are a 370–390°C nozzle, a 150°C plate and roughly a 90°C chamber, with drying at 120°C for 8 hours. If your machine cannot hold those numbers, the honest answer is that it is not ready, and a service bureau or a machine with in-situ crystallisation is the shorter path.

On the question of printing PEEK at 350°C: the melt point is about 343°C, so 350°C is barely above it. You will get poor flow, stringing, and weak layer bonding. One converter described PEEK as gummy at 400°C with flow improving as the temperature rose toward 430°C, and successful prints on that machine ran at 385–400°C nozzle with a 110–150°C bed.

2. Dry and prepare the filament

Moisture is the quiet cause of most PEEK problems. Wet filament pops and spits in the nozzle, the vapour bubbles out mid-layer, stringing gets worse, and the part looks scorched even when it was not. Published guidance for AM 450 FIL puts the ceiling at 0.020% moisture.

Dry in a dedicated high-temperature dryer using the supplier’s cycle: 120°C for 5 hours is the published figure for that grade, and 120°C for 8 hours for ULTEM AM1010F. A food dehydrator modified for temperature control works for a lot of hobbyists, but check that it holds temperature accurately at the top of the cycle rather than cycling wildly.

PEI is more forgiving than PEEK, which is another reason to start there, but dry it anyway. Handle dried filament with dry gloves and transfer it straight into a sealed dry box with desiccant. Print from the dry box or a sealed spool box; every minute of open-air time in a humid room works against you. Condition filament that has been stored open before its first print — moisture uptake from storage is separate from moisture absorbed during a print.

3. Prepare the extruder and heated chamber

Start with a clean, unblocked nozzle. Cold-pull maintenance on a regular machine habitually leaves a plug of plastic in the melt zone, and at 400°C that plug deforms and presses upward into the heater block, killing the print hours in. If the nozzle has PEEK frozen in it, the recovery procedure is in the troubleshooting section below.

Confirm the filament path contains no PTFE above the heat break, that the heat break is seated and the fan path is unobstructed, and that the chamber heater is holding setpoint before you load material.

Load the filament and purge at roughly 380–400°C for PEEK, or 370–385°C for PEI, directly into the build plate or a scrap block rather than into the air. Purge volume is substantial for these materials and the melt strength of PEEK makes the purge stringy and long. A purge tower is safer than letting it drip.

Select the build surface with care, because this is a real trade-off and there is no best option in the abstract.

Build surfaceWhere it worksWhere it fails
PEI sheet (ULTEM or powder-coated)Very strong adhesion, survives high bed temperatures, reusable with proper releaseAdhesion can exceed the part’s own strength and tear corners off during removal, as one converter found after switching from polyimide tape
Polyimide (Kapton) tapeConsistent first-layer grip at high temperature, forgiving, easy to replace when it degradesWrinkles and bubbles appear over time; multiple layers of tape lose grip; replaces frequently
PEI tapeMiddle ground: better release than a rigid PEI sheet, less maintenance than taped KaptonRides up at the edges on long prints if not seated carefully
Adhesive powder or glueWorks on bare aluminium, releases cleanly, no plastic surface to deformLoses strength at extended high bed temperatures; needs reapplication often

Mask the bed surface away from the part footprint. That single step prevents molten polymer from fusing to the surface and makes removal far less violent.

4. Set a material-specific temperature profile for PEEK and PEI

Set from the supplier’s validated profile for your grade, then adjust one variable at a time. Treat the table below as a starting range, not a guarantee — every chamber, every filament batch, and every part geometry behaves a little differently.

Set a material-specific temperature profile for PEEK and PEI
SettingPEEK starting rangePEI starting range
Nozzle temperature380–430°C370–390°C
Bed temperature120–160°C, above the glass transition of about 143°C110–150°C
Chamber temperature80–150°C, higher for larger or stronger parts80–100°C
Initial layer height0.3–0.4mm0.2–0.3mm
First layer speed20–30mm/s20–30mm/s
Maximum print speed40–60mm/s; some machines accept up to 180mm/s but layer bonding degrades40–55mm/s
Part cooling fanOff or minimal, usually 0–20%Off or minimal
Drying120°C for about 5 hours, below 0.020% moisture120°C for about 8 hours
Annealing170–180°C cycle per supplier guidanceOptional; properties do not depend on it

Speed matters more than most slicer defaults allow. The frequently quoted optimum for high-performance polymers is around 55mm/s, with 180–250mm/s treated as a machine maximum rather than a working speed. If you print hot and fast, expect stringing and a weak Z axis.

Add a brim of 10 to 30 outlines for first-layer adhesion on these materials. The extra plastic is trivially removable; a lifted bottom layer is not.

Keep every layer above the glass transition temperature for the whole print. That means no part cooling fan, no draft in the chamber, and a chamber that has soaked for 20 to 30 minutes after reaching setpoint before the first layer starts. Cold corners in a large chamber produce warped corners, and warped corners lift off the plate mid-print.

Design for Z. Inter-layer adhesion is the weakest axis in a printed part regardless of material, so orient load paths along the XY plane, avoid tall thin sections, and split tall parts into bonded sections if stiffness in Z matters.

5. Calibrate, test, and complete the print

Print a coupon first — a 20–40mm cube with an optional wall — not your actual part. On this coupon you can check extrusion consistency, layer bonding, corner warp, and first-layer adhesion in a fraction of the time a failure would cost you.

To break a coupon by hand, a healthy PEEK or PEI part resists layer separation and snaps in the bulk of the section. If it splits cleanly at a layer line, your chamber is too cold, your speed too high, or your filament wet.

Then check the printed part for the things that only show up in geometry: stringing between features, extrusion gaps, cracking at corners, brim adhesion, and support contact marks. Support material must be the same polymer as the part — there is no soluble support that dissolves reliably in these materials — which means designing support interfaces that release with minimal mechanical force.

Let the machine cool on its own before you touch anything. Do not open the chamber the moment the print finishes. Cracking and warping during cooldown is the single most avoidable failure in this workflow, and it comes from people in a hurry with a part cutter.

Remove the part while the plate is still slightly warm rather than fully cold, and support it with both hands. Fully cold PEEK is more brittle, and a cold snap can crack a corner that was fine.

6. Anneal and remove the part safely

Annealing is how you raise PEEK’s crystallinity, which raises stiffness, dimensional stability, and chemical resistance. It is optional for PEI because amorphous PEI has no crystallinity to develop.

The published guidance for AM 450 FIL points to a cycle in the 170–180°C range. Ramp slowly, hold, and cool slowly. Fast ramps create internal stress and a fast cool-down creates distortion — a part that comes out of the oven visibly warped or dished has usually been through a cycle that was too quick rather than one that was too hot.

Support the part during the cycle. Large parts need a fixture or sacrificial support because gravity plus soft polymer at 170°C adds up. Expect dimensional change and plan your tolerances accordingly, especially for anything with a fit. Annealing also unifies colour and surface appearance across the part, because different regions of the print reached different crystallinity.

Always follow your polymer supplier’s cycle. This is the one step where improvising with a kitchen oven is a good way to ruin a part you spent hours on.

7. Common mistakes to avoid

Most of these are covered in the troubleshooting table below, but three come up often enough to name here.

Printing with the part cooling fan on, or leaving the chamber door ajar, breaks the entire premise of the process. The chamber exists to keep every layer hot; defeating it means printing a weaker, more warp-prone part at a much higher material cost.

Over-speeding to save time is the second. At 40 to 60mm/s these polymers already take hours on a small part.

Skipping filament drying is the third. It costs a few hours and it fixes problems people spend days chasing.

Common Mistakes: Troubleshooting PEEK and PEI Prints

Each row below maps a symptom to its most likely cause, the corrective action, and the observation that tells you the fix worked. Change one thing at a time, because these materials punish simultaneous changes.

SymptomLikely causeCorrective actionConfirming observation
Bottom layer lifts or corners curlBed below the glass transition temperature of the polymer, or too little brimRaise bed toward 150°C, add 10–30 brim outlines, verify surface is clean and levelFirst layer stays bonded through the full print; corners sit flat at cooldown
Filament deforms and pushes up into the nozzle blockCold pull left a plug in the melt zone that softened and travelled upwardClear the nozzle mechanically, then purge at temperature before every loadClean extrusion resumes and flow is steady for the whole spool
Nozzle clogs repeatedly mid-printDamp filament, excessive retraction, or a cold zone in the melt breakDry filament properly, disable retraction or reduce it, verify no PTFE above the heat breakLongest uninterrupted run matches your spool estimate
Heavy stringingTemperature below the polymer’s practical melt window, which is what makes PEEK look gummy at 400°CRaise nozzle in 5°C steps toward the upper end of the supplier’s rangeStrings shorten then stop; surface texture tightens
Poor layer bonding, part splits at a layer lineChamber too cold, print speed too high, or moistureRaise chamber setpoint, drop speed to 40–55mm/s, verify moisture is under 0.020%Break test fails in the bulk of the section rather than along a layer
Corners warped, centre flatTemperature gradients in the chamber and insufficient soak timeExtend heat soak to 30 minutes, improve chamber insulation, reduce part sizeCorners hold position from layer to layer
Part cracks after removal or after annealingCool-down shock, or an anneal ramp that was too fastLet the chamber cool fully, support the part during handling, slow the anneal rampPart stays intact through a full cooldown and anneal cycle
Corner of the part tears off on removalPEI sheet adhesion exceeded the part’s own strengthUse polyimide tape or powder glue, or score and break the sheet around the partPart releases without force and without material loss
Stringy, discoloured, brittle extrusionNozzle temperature above the supplier’s melt window, or air gap drawing fumes into the meltDrop nozzle temperature, close the chamber fully, add an air gap before the nozzleExtrusion colour returns to normal and remains consistent
Extrusion gaps on thin wallsVolumetric flow exceeded, or moisture-driven poppingLower volumetric limit in the slicer, increase layer height slightly, re-dry filamentWalls close consistently across the full height
Firmware temperature lock or timeout errorHeater cartridge wattage too low, or a sensor failing above 350°CFit a higher-wattage cartridge, replace PT100 with a K-type thermocouple and matching amplifierNozzle holds setpoint for the full print without a lock error
Support removal cracks the partSame-material support with no designed release geometryRedesign interfaces with draft and a small sacrificial gap, or machine the interface after printingSupports release with hand tools only

When the nozzle is clogged and nothing comes out

This is the top complaint in the PEEK community and the reason people abandon the material. Soaking a clogged nozzle in acetone does nothing useful against PEEK, because PEEK resists solvents and the plug is fused to the metal.

The practical sequence: heat the hotend to its maximum validated temperature and let it soak so the plug softens, then push filament through under pressure. If that fails, remove the nozzle hot and let it cool, then push the solidified plug out mechanically from the wide end with a nozzle-cleaning needle or an inert rod sized to the bore. Ice water on the nozzle tip after removal helps the plug contract and release. Replace the nozzle if the bore is damaged, because a scored or deformed orifice causes under-extrusion long after the clog is gone.

People running filled grades such as PEEK-CF or PEKK report the same clogging behaviour as unfilled PEEK, so the same procedure applies.

Safety and fumes

Run extraction before the heater comes on, and keep it running through cooldown. Forum threads on HT conversions consistently treat a fume hood as the baseline practice, not an upgrade.

Both polymers degrade before they fail catastrophically, and degraded extrusion looks different from healthy extrusion: stringier, discoloured, brittle. That is your early warning that dwell time is too long or temperature too high.

On regulated parts, be clear about boundaries. FFF output is not automatically suitable for food contact, medical implant, or aerospace-certified use even when the polymer grade itself carries the approval. Chemical and food contact statements usually assume injection moulded or vendor-qualified parts. For anything regulated, print via a service bureau that can document the process, or qualify your own machine properly.

Frequently Asked Questions

Can a regular FDM printer print PEEK or PEI?

Not reliably. You need an all-metal hotend rated to 500°C, an actively heated chamber, a bed reaching at least 120°C, and an enclosed frame. Community conversions of consumer printers did produce successful PEEK prints at 385-400°C nozzle with a 110-150°C bed, but only after fitting a 40W heater cartridge, insulating the heat block, and replacing a PT100 sensor that failed at those temperatures. Most people end up spending more than a used entry-level HT machine costs.

What is the difference between PEEK and PEI for 3D printing?

PEEK is semi-crystalline, so its strength depends on how much crystalline structure develops during printing and post-print annealing. PEI is amorphous and reaches its properties as printed. In practice PEEK is harder to print: higher nozzle temperatures, hotter chamber, tighter moisture control, and crystallinity distortion risk during annealing. PEI 9085 prints more predictably and is the sensible starting point on a machine with an 80°C chamber.

Why does PEEK or PEI filament need to be dried before printing?

Both polymers absorb moisture from the air, and steam forms inside the melt zone during extrusion. That steam pops at the nozzle, causes stringing, leaves voids between layers, and makes parts brittle. Published guidance for VICTREX AM 450 FIL puts the ceiling at 0.020% moisture, with drying at 120°C for around 5 hours. ULTEM AM1010F is published at 120°C for 8 hours. Print straight from a sealed dry box and wear dry gloves when handling.

What temperature should a high-temperature chamber be for PEEK and PEI?

Around 80°C is the practical floor for reliable small PEI parts and roughly the ceiling for most converted consumer machines. Around 100°C unlocks larger PEI parts and better PEEK crystallisation. At about 150°C, near the glass transition region, PEEK approaches in-situ crystallisation and larger parts become practical. Around 200°C enables full crystallisation during the print. Keep every layer above the glass transition for the whole print and soak the chamber for 20 to 30 minutes first.

How can I reduce fumes and warping when printing PEEK or PEI?

For fumes, start extraction before the heater switches on, keep the chamber sealed, add an air gap before the nozzle, and lower the nozzle temperature if you see discoloured or stringy extrusion. Fume extraction and a sealed chamber are standard practice in the high-temperature printing community, not optional extras. For warping, heat soak the chamber for 20 to 30 minutes, raise the bed above the glass transition temperature, add 10 to 30 brim outlines, orient load paths in the XY plane, and keep the part small enough for your chamber tier.

Conclusion

Four things decide whether a high-temperature print works, and the order matters. Identify the exact polymer and grade from the spool, then verify your machine can hold nozzle, bed, and chamber temperatures for the supplier’s validated profile. Dry the filament to the published moisture ceiling and print from a sealed box. Run a small coupon, check layer bonding and corner warp, and only then commit to a full part — with an anneal cycle at the end if you are working with PEEK and need the extra crystallinity.

Leave a Comment