Organ-on-chip platforms compared: Emulate, MIMETAS, CN Bio and six others

The honest headline for this comparison is that the most decision-relevant column — what a well costs you — is empty for every platform on the market. None of the eight vendors below publishes consumable pricing, instrument pricing or a cost per data point. That is not an oversight in our research; it is the market structure, and it is the single largest obstacle to rational platform selection.

What can be compared from published sources is architecture: throughput, how fluid moves, whether tissue sits on a membrane or free in a gel, what imaging the format supports, how closed the consumable ecosystem is, and which tissue models the vendor has actually published. Those determine whether a platform can answer your question at all, which is the question that comes before price.

The comparison

The four architectural choices that actually decide fit

Platform marketing converges on the same vocabulary, so the useful discriminators are structural.

Membrane or no membrane

Emulate’s architecture is two channels separated by a porous membrane, which is the canonical organ-chip design and the right one when you need a defined apical and basolateral compartment with a physical barrier you can characterise. MIMETAS took the opposite decision: the OrganoPlate uses PhaseGuide capillary pressure barriers to pattern ECM and channels without any membrane at all, so an epithelial tubule sits directly against a collagen matrix.

This is not a quality difference, it is a question difference. A membrane gives you a well-defined transport surface and a familiar mental model inherited from Transwell work. Membrane-free gives you direct cell-to-matrix contact and unobstructed imaging through the tissue interface, at the cost of a less crisply defined geometry.

If your readout is transepithelial transport with a clean area term, membrane. If your readout is angiogenic sprouting, migration into matrix, or anything that needs cells to invade rather than sit, membrane-free.

Pumped or gravity-driven

CN Bio, Emulate and TissUse all move fluid actively. MIMETAS moves it by putting the plate on a rocker and letting gravity do the work through the OrganoFlow, with the vendor emphasising that all you need is a plate, a pipette and the rocker.

Active perfusion gives controlled, tunable shear and enables recirculating multi-organ linkage — CN Bio’s gut/liver configuration and TissUse’s circulating multi-organ chips both depend on it. Gravity-driven perfusion gives up shear control and buys back enormous operational simplicity: no pumps, no tubing, no bubble management, no per-chip fluidic connections, and a footprint that drops straight into plate-handling automation.

For most laboratories the operational difference dominates. Bubble management and tubing failure are the two things that most often ruin a chip experiment, and one of these architectures does not have either failure mode.

Material, and therefore absorption

This is where the material discussion becomes commercially visible. Emulate sells the Chip-R1 Rigid Chip explicitly on a low-drug-absorption profile for ADME and toxicology work, alongside its elastomeric Chip-S1. MIMETAS states non-absorbent materials as a platform property and builds on 150 µm microscope-grade glass.

Both are direct commercial responses to a documented problem. Kemas and colleagues fabricated identical microdevices from eight polymers and found absorption differing by more than 1000-fold, with PDMS the most absorptive; more than 99% of chlorpromazine partitioned into the PDMS bulk within 24 hours, and correcting for absorption moved the in vitro toxic concentration into agreement with clinical plasma values. If your assay involves a compound with logP above roughly 2, the chip material is not a detail. See the material guide for the full data.

Ready-to-use tissue or bring your own cells

A genuine axis that cuts across the platform question. MIMETAS sells OrganoReady plates with perfused tissues already formed by their staff. InSphero ships ready-to-use microtissues on a published production schedule. Emulate sells bioKits bundling chips with pre-qualified cells and activation reagents. Altis sells RepliGut Kits.

Buying tissue rather than culturing it removes the single largest source of experimental variance and the single largest labour cost, and replaces both with a delivery schedule you do not control and a cell source you did not choose. For a group with no dedicated culture staff, it is often the only workable route. For a group with a specific donor, genotype or disease line, it is not available at all.

Throughput numbers are not experiment counts

Every throughput figure in the table is the vendor’s chip or tissue count. Translating that into experimental power requires two divisions that vendors do not do for you.

Replicates. Chip-to-chip variability in perfused 3D culture is generally higher than well-to-well variability in a 2D plate. If you need four technical replicates per condition, a 64-chip plate is a 16-condition plate.

Controls. Vehicle, positive control and often a barrier-integrity control per plate. On a 40-chip format that is a meaningful fraction of the plate before any test article is applied.

A 96-chip format at four replicates and three controls per plate is roughly a 21-condition experiment. That is the number to compare against a 384-well plate assay, and when you do, the throughput argument for organ-chips looks much weaker — which is correct. Organ-chips are not a throughput technology. They are a physiological-relevance technology with a throughput cost, and the honest comparison is against the assay they replace, not against the plate count.

The pricing vacuum, and how to work around it

Since no vendor publishes consumable cost, you cannot compare platforms on cost per data point from public information. What you can do is force the comparison into the quote request.

Ask each vendor for the same five numbers:

1. Instrument cost, including any required accessories — culture module, rocker, controller, software licence, and whether the licence is perpetual or annual.
2. Consumable cost per plate at three annual volumes — say 10, 50 and 200 plates per year. The volume discount curve differs sharply between vendors and is invisible at a single quantity.
3. Cost of the cells, if the platform requires or strongly prefers vendor-supplied cells. On several of these platforms the cells cost more than the consumable.
4. Ready-to-use premium, where offered — the delta between a bare plate and a plate with tissues already formed, which prices the vendor’s own culture labour and tells you a great deal about the real difficulty of the workflow.
5. Three-year total on your actual planned experiment count, including service contract.

Then divide by usable conditions, not by chips. That number is comparable across platforms; the per-chip price is not.

Which platform for which question

A blunt mapping, offered as a starting point rather than a recommendation:

• Barrier function and transport with defined geometry, plus mechanical strain: Emulate. The membrane architecture and Chip-S1 stretch capability are purpose-built for it, and Chip-R1 exists for when absorption would otherwise compromise the ADME readout.
• Perfused tubules, angiogenesis, migration, and high plate counts with minimal fluidic complexity: MIMETAS. Membrane-free, pumpless, plate-footprint, glass-bottomed, and the widest published model range in the set.
• Liver-centric ADME and toxicology, and multi-organ linkage: CN Bio. The published application focus — MASLD/MASH, DILI, absorption, metabolism, bioavailability, oligonucleotide delivery — is narrower and deeper than the others, and they will run it as a service if you would rather not own the box.
• Multi-organ circulating systems in a configurable chip: TissUse. The HUMIMIC line is built around configurable multi-compartment setups rather than a fixed format.
• Scaffold-free 3D biology with reliable supply and no culture burden: InSphero. Ready-to-use microtissues on a published production schedule is the strongest supply-side proposition in the category.
• Human intestinal epithelium specifically: Altis. RepliGut is planar and static rather than perfused, which for many gut questions is the appropriate and cheaper architecture.
• Screening scale delivered as a service with image-based analysis: Xellar. Platform access is the product rather than the hardware.

What none of them do

No platform in this comparison supports contractile skeletal muscle with force readout. That capability sits in a different device category entirely — post-and-pillar casting devices from eNUVIO and Curi Bio — covered on the 3D muscle culture devices page. If someone tells you an organ-on-chip platform will give you engineered muscle force data, check which of the two categories they are actually selling.

Equally, none of these vendors sells custom chip geometry. Every one is a fixed-format consumable ecosystem. If you need a geometry that does not exist, that is a fabrication project — see custom microfluidic device fabrication.

How we can help

We run the same five-question pricing enquiry across these vendors simultaneously and normalise the answers to a cost per usable condition over a three-year horizon. We hold no reseller relationship with any of them, which means we can tell you when the answer is a Transwell plate and a rocker. Send us the assay, the compound properties, the required conditions per year and whether you have culture staff.