3D Skeletal Muscle Models Compared: Myobundles, Micropillars and Micropatterns (2026)

The short answer: five distinct 3D and quasi-3D skeletal muscle model architectures are commercially reachable in 2026, and they answer different questions. Two publish a price. One sells finished tissue. One is a service. And the force numbers they produce are not comparable to each other, because post and stretcher stiffness changes the measured force by nearly a factor of two on the same tissue.

If you only take one thing from this page: normalise for anchor stiffness before comparing any two force figures, including two from the same platform.

The comparison matrix

Model	Supplier / origin	Cell source	Device and format	Force readout	Maturation time	Throughput	Consumable cost per tissue
Mantarray EMT	Curi Bio (Seattle)	iPSC-derived myogenic progenitors or primary human myoblasts; vendor sells both, and disease lines including DMD, DM1 and XLMTM	24-well casting plate with paired flexible posts; Mini Plates reduce tissue volume up to 60% and cells up to 50%	Magnetic sensing — magnet in the flexible post tip over giant magnetoresistive sensors; real-time, label-free, all 24 wells in parallel	Published force plateau at day 10 for iPSC-derived EMT; primary myoblast tissues still rising at day 14	24 tissues per plate, parallel acquisition	Quote only
OMEGA-MP	eNUVIO (Montréal)	Myogenic progenitors from primary, iPSC or immortalised sources, resuspended in an ECM-derived hydrogel and seeded by the customer	4 PDMS devices, 2 chambers each, fits 12-well plates; 21.25 mm diameter; chamber ~0.25 cm²; 25–30 µL seeding volume; pillars 2.7 mm tall, 1 mm wide	Optical post deflection, measured with your own imaging. Open-top chambers give full access for electrophysiological stimulation or recording	Vendor states 3D muscle matures more rapidly and completely than monolayer, and is stable for longer-term culture; no day number published	8 experiments per kit	$55.00 ($440 for 8 experiments)
OMEGA96-MP	eNUVIO	Same	Single 96-well standard microplate, pre-bonded high-transmissive PDMS bottom, two micropillars per well	Optical; conforms to the 96-well microplate standard so it works with plate washers, readers, high-content imagers and plate hotels	Same	96 experiments per plate	Quote only
Myobundle	Duke (Bursac lab) protocol — academic, not a product	Primary human myogenic cells from needle biopsy or surgical waste, and commercially available myoblasts	Fabricate yourself: cells in fibrinogen/Matrigel on laser-cut Cerex frames (9.2 × 9.5 mm outer, 6.8 × 8.3 mm inner) within PDMS moulds cast from Teflon masters, at 15 × 10⁶ cells/mL, 7.5 × 10⁵ cells per bundle	Custom force transducer; the published work also demonstrates GCaMP6 calcium imaging correlating with contractile force for non-invasive tracking	Twitch and tetanus forces increase over time with significant enhancement at 4 weeks versus 1 week	Whatever you build	No product to buy. Consumables only, plus mould fabrication
myrPlate + myrTissue-Skeletal	myriamed (Göttingen)	iPSC-derived myoblasts (hypaxial origin) — sold as cells, as a plate, or as finished tissue	48-well SBS plate, two flexible stretchers per well, tissue in ring format; six stiffness grades TM4-MED to TM9-MED	Optical — stretchers contain a fluorescent dye for optical tracing of developed force	Finished tissue: 7–11 weeks order-related manufacturing	48 wells per plate	Plate: €5.63 (€270/48 wells; €4.58 at the 10-plate price). Finished tissue: €1,400 / €1,080 / €667 for 1 / 5 / 12
MyoScreen	CYTOO (Grenoble)	Patient-derived primary myoblasts from an established patient cell bank plus biobank collaborations. Named disease coverage: DMD, DM1, DM2, SMA, FSHD, Pompe, muscle wasting, cardiometabolic	Micropatterned plates — this is aligned 2D, not 3D. Micropatterning constrains myotube geometry to produce uniform linear myotubes	Not force. High-content imaging readouts: fusion index, sarcomeric striation, AChR clustering in pretzel-like structures, plus machine-learning phenotypic classification	Vendor states maturation is significantly greater and faster on MyoScreen plates than conventional culture	High-throughput by design; automation-ready	Quote only — CYTOOchips and CYTOOplates sold as products; screening runs as a service

The finding that invalidates most force comparisons

Post stiffness changes measured force by nearly a factor of two on identical tissue, and two of five stiffnesses tested produced no usable data at all.

From the Mantarray validation study, iPSC-derived engineered muscle tissue at day 10:

Post stiffness	Average twitch force at day 10
1.92 N/m	1156.3 ± 129.1 µN
1.44 N/m	919.7 ± 185.3 µN
0.96 N/m	624.0 ± 154.6 µN
0.48 N/m	No usable data — posts bent under passive tension to the point where active contraction could not be resolved
0.16 N/m	No usable data — same failure

The difference between the stiffest and softest usable posts was statistically significant (p = 0.04). The authors interpret softer posts producing smaller forces as increased resistance to contraction promoting skeletal muscle function, consistent with in vivo loading.

Three consequences.

A twitch force in µN is meaningless without the anchor stiffness. Any vendor comparison quoting absolute force without stating stiffness is not a comparison.
Too soft fails outright, not gracefully. If you commission a custom device or select a stiffness grade, it needs a value and a tolerance. This is why myriamed’s €330 gradient plate — one row of each stiffness from TM4-MED to TM9-MED — is the correct first purchase for anyone new to a stretcher-based platform. It costs €330 to avoid casting a whole experiment into the unresolvable region.
Stiffness is a maturation lever. If a construct underperforms, stiffening the anchors is cheaper than changing cell source.

Force numbers, with their conditions attached

Platform	Cell source	Metric	Value	Conditions	Source
Mantarray	iPSC-derived myogenic	Average twitch	385.8 ± 51.1 µN	Day 7, 2% horse serum throughout, 1.92 N/m posts	Smith et al. 2022
Mantarray	iPSC-derived myogenic	Average twitch	1156.3 ± 129.1 µN	Day 10, same conditions	Smith et al. 2022
Mantarray	iPSC-derived myogenic	Average twitch	943.1 ± 162.1 µN	Day 14 — force plateaued after day 10, no discernible improvement thereafter	Smith et al. 2022
Mantarray	Primary human myoblasts (Lonza CC-2580)	Average twitch	306.5 ± 22.2 / 468.8 ± 37.5 / 568.0 ± 103.7 µN	Days 7 / 10 / 14 — still rising at day 14	Smith et al. 2022
Mantarray	Primary human myoblasts	Average tetanic	740.8 ± 91.3 / 1160.5 ± 118.8 / 1387.7 ± 275.7 µN	Days 7 / 10 / 14	Smith et al. 2022
Duke myobundle	Primary human myogenic cells, ten sources: nine donor muscle samples plus one commercial myoblast source	Specific force, twitch	2.1 ± 0.9 mN/mm²	Fibrin/Matrigel on Cerex frames, 7.5 × 10⁵ cells per bundle	Madden et al. 2015
Duke myobundle	Same ten sources	Specific force, tetanus	7.0 ± 2.2 mN/mm²	Same. Average tetanus-to-twitch ratio 3.5 ± 0.8	Madden et al. 2015

The two contexts that make those numbers interpretable:

The Duke tetanic specific force was similar to values measured in fetal human muscle and an order of magnitude lower than values reported for adult muscle. The authors say so directly. That is the honest maturity position of engineered human skeletal muscle, and any vendor implying adult-equivalent force should be asked for the specific force in mN/mm², not the absolute force in µN.

The Duke result also holds across ten independent cell sources — nine donors plus one commercial myoblast line — which is a robustness demonstration no single-source vendor figure can match.

Absolute force versus specific force

These two are constantly conflated and they answer different questions.

Absolute force (µN or mN) is what the tissue pulls with. It depends on cross-section. It is the number you need if you are building an actuator that must move something.

Specific force (mN/mm²) is force normalised to cross-sectional area. It is the number you need if you are asking whether the muscle is any good, because it is comparable across constructs of different sizes and against native tissue.

A platform quoting only absolute force is telling you about its geometry as much as about its biology. Ask for both, plus the cross-section used.

Maturation time: the timelines that actually differ

Route	Time to usable force	What sets the clock
Mantarray, iPSC-derived	Force plateau at day 10. No reproducible improvement after that, so subsequent analysis in the validation study was performed at that timepoint	Fusion and maturation in 2% horse serum. The study found 2% horse serum throughout produced the greatest force development; 2% KSR and 15% FBS both performed notably worse
Mantarray, primary myoblasts	Still rising at day 14	Slower, more gradual force progression than iPSC-derived tissues
Duke myobundle	Significant enhancement at 4 weeks versus 1 week	Longer maturation window, larger myofiber diameter, improved calcium handling
myriamed finished tissue	7–11 weeks order-related manufacturing before it reaches you	Full manufacturing cycle: cells, casting, maturation, release
eNUVIO OMEGA-MP	Not published as a day number	Your protocol, your cells

The medium finding from the Mantarray work is worth isolating because it is free to act on: maintaining tissues in 2% horse serum throughout culture produced the greatest force development of the three formulations tested. Under 2% KSR, average twitch forces at days 7, 10 and 14 were 413.0, 459.3 and 100.2 µN respectively — the day-14 collapse being the striking part. If your constructs are weakening rather than strengthening, the medium is the first thing to examine, not the cell source.

Cost per data point, calculated where it is honest to do so

Only two platforms publish enough to compute anything. Here is what can be computed, and what it excludes.

Platform	Public price	Tissues	Device cost per tissue
eNUVIO OMEGA-MP	$440.00	8 experiments	$55.00
myriamed myrPlate Uniform, 1 plate	€270.00	48 wells	€5.63
myriamed myrPlate Uniform, 10 plates	€2,200.00	480 wells	€4.58
myriamed myrPlate Gradient, 1 plate	€330.00	48 wells	€6.88
myriamed myrTissue-Skeletal, finished	€1,400 / €5,400 / €8,000	1 / 5 / 12 tissues	€1,400 / €1,080 / €667 per tissue, cells and labour included
Curi Bio Mantarray plate kit	Quote only	24	—
eNUVIO OMEGA96-MP	Quote only	96	—
CYTOO MyoScreen	Quote only	—	—

Read those numbers carefully, because the spread is misleading. The myrPlate at €5.63 per well and the OMEGA-MP at $55 per experiment are not competing on the same axis: the myrPlate is a 48-well plate of small ring tissues, the OMEGA-MP is four separate PDMS devices with larger chambers, and the tissue volumes and cell requirements differ accordingly. Cost per well is not cost per data point when the wells are different sizes.

What these figures exclude, in rough order of magnitude:

Cells. The dominant cost. bit.bio’s ioSkeletal Myocytes are 1,198 for a 2-vial pack of more than 2.5 million viable cells each, or 2,398 for 2 × 5 million; DMD exon-deletion disease lines run 1,558 to 3,118. myriamed’s myrCell-Skeletal is €750 per million. At the Duke protocol’s 7.5 × 10⁵ cells per bundle, the cell input alone is a multiple of any device cost on this page.
Matrix. Fibrinogen, Matrigel or an equivalent basement membrane extract. Lot variability here is a real source of tissue-to-tissue variation.
Media across a 10-day to 4-week culture, at whatever formulation your force data depends on.
Stimulation hardware, if you want maturation.
The readout instrument, or an imaging pipeline and the labour to run it.
Several weeks of skilled culture labour per run.

The honest conclusion on cost: the device is never the number that decides the programme. A first experiment on OMEGA-MP devices is a $440 purchase inside a several-thousand-dollar experiment. The levers that actually move total cost are cells per tissue — which is why Curi Bio’s Mini Plate claim of up to 50 percent fewer cells matters more than any plate price — and whether you buy tissue or build it.

Where each model is strongest

Being fair to each of these requires saying what it is actually best at, because they are not competitors so much as different tools.

Curi Bio Mantarray is strongest on automated parallel force acquisition and on published validation. It is the platform with a peer-reviewed characterisation study covering post stiffness, media formulation, cell source and stimulation frequency, which is more than any other option here offers. Magnetic sensing across 24 wells in real time without imaging is a genuine capability advantage for a screening programme. The cost is instrument commitment and quote-only consumables.

eNUVIO OMEGA-MP is strongest on price transparency, openness and custom routes. It is the only 3D muscle device in the market with a hard public price, its chambers are open-top for electrode access, it is fully transparent for imaging, ships sterile with a two-year shelf life, and eNUVIO runs a cleanroom microfabrication service that will build a modified mould from scratch. For a group that owns a microscope and wants to try 3D muscle without a capital decision, it is the obvious first purchase.

The Duke myobundle protocol is strongest on scientific pedigree and cross-donor robustness. It is not a product and you cannot buy it, but it is the most thoroughly published human myobundle methodology, validated across ten cell sources with specific force referenced to native fetal muscle and drug responses that mirror clinical outcomes. If you are building capability rather than buying a platform, this is the literature to build from.

myriamed is strongest on the buy-versus-build choice. It is the only supplier here that will sell you the cells, the plate, or a finished living tissue, at published prices, across all three. The gradient plate is the best-designed onboarding product in the category. The trade is a seven-to-eleven week lead time on finished tissue.

CYTOO MyoScreen is strongest on patient-derived disease modelling at throughput. It is deliberately not a force platform and should not be compared as one. What it offers instead is an established patient cell bank across eight named disease areas, micropatterning that removes the myotube heterogeneity that makes conventional cultures incompatible with high-content analysis, and a quantitative imaging and machine-learning analysis layer. If your endpoint is a phenotypic screen on patient cells rather than a force measurement, it is a different and better answer than anything else on this page.

Choosing, honestly

You need a force number and you need it repeatedly. Mantarray, and budget for the instrument and three years of consumables.
You need a force number once, cheaply, and you own a microscope. eNUVIO OMEGA-MP at $55 per experiment, or a myrPlate with the gradient plate first.
You need muscle tissue but do not have culture capability. myrTissue-Skeletal, and start planning eleven weeks out.
You need a phenotypic screen on patient-derived cells. CYTOO MyoScreen, and stop asking about force.
You need a specific geometry that does not exist. Nobody sells it. eNUVIO’s custom microfabrication service is the realistic first call, and the account of what that costs and how long it takes is in engineered skeletal muscle tissue.
You need adult-equivalent specific force. Not available from any commercial platform. Published engineered human muscle sits an order of magnitude below adult native tissue, and the honest answer is to design around that rather than shop for it.

Where to go next

3D muscle culture devices goes deeper on the hardware — post geometry, tissue volume, stimulation access and the custom fabrication route. Engineered skeletal muscle tissue covers who will quote a finished force-specified construct and what that actually involves. Human skeletal myoblasts covers the cell input, which is the cost that dominates every route on this page.

How we can help

The comparison this page cannot complete is the one that needs quote-only numbers: Mantarray consumables, the OMEGA96-MP plate, CYTOO screening packages. We will get those and put them beside the published prices on a cost-per-tissue basis with the cell input included, which is the only version of this comparison that reflects what you will actually spend.

We will also ask each supplier the question that decides transferability and that none of them publish: what anchor stiffness their headline force figure was measured at, and what the specific force in mN/mm² was. Tell us the tissue count, the readout, the cell source and whether you already own an instrument.

3D skeletal muscle models compared: device, cell source, force readout and cost per data point