biacore 8k+ vs octet r8 vs carterra lsa · biacore 8k+
Biacore 8K+ vs Octet R8 vs Carterra LSA: Kinetics Compared
November 2, 2026
29 min read
Vendor-neutral 2026 comparison of Biacore 8K+, Octet R8 and Carterra LSA: architecture, fitting and review, data export, 2020 study limits, quote-only pricing and a pilot plus cost-per-accepted-interaction model.
- 01Match the platform to the assay: BLI supports fast titer and screening, flow-based SPR supports kinetic confirmation, and array SPR suits large antibody binning panels.
- 02The direct SPR study tested the older Biacore 8K against LSA. Older Octet studies cannot establish R8 accuracy, and no R8-specific independent kinetics benchmark was located.
- 03Treat raw-data export and reconstruction in the target ELN or registry as acceptance gates. Software modules, access controls and validation depend on the purchased configuration.
- 04Use a blinded same-panel pilot with thresholds set before review, then compare cost per accepted interaction. Vendor throughput claims use different units and cannot be ranked directly.
- 05Name the exact model, software edition and licences in every quote. All three vendors have newer family members, and results from older models may not transfer.
Executive Summary
This report compares three label-free binding platforms that antibody teams often shortlist together: the Cytiva Biacore 8K+ (flow-based surface plasmon resonance, SPR), the Sartorius Octet R8 (dip-and-read biolayer interferometry, BLI) and the Carterra LSA (array-based high-throughput SPR). It reflects vendor documents and papers retrieved on 1 to 2 November 2026. No instrument was tested hands-on, and no universal winner emerges. The decision turns on assay fit (titer, kinetic confirmation or large binning panels), review burden and data portability, not on channel counts or headline throughput.
Procurement starts with a naming trap. Cytiva announced the Biacore 8S and 8S+ as the next generation of the Biacore 8 series in May 2026 ([1]), Sartorius markets the Octet R8e as its latest BLI system, and Carterra now lists LSAXT, Ultra and Vega beside the LSA, with Vega announced as a 48-channel platform in February 2026 ([2]). A quote must therefore name the exact model, software edition and licence list. Vendor throughput claims use different units (samples per tray, samples per plate, analytes per 24 hours) and cannot be ranked against each other.
The evidence base is narrower than the marketing. The only direct SPR comparison of two of the platforms (Biacore 8K and LSA, 2020) found that 20 of 23 Fab affinities agreed within 1.5-fold ([3]), and the study used the older 8K, skilled users and defined reagents. Independent BLI benchmarks come from earlier Octet generations; one reported that BLI underestimated the affinity of low-picomolar antibodies ([4]). Solution-phase checks are slow and agree only partly with SPR arrays: a 48-antibody comparison reported 54% correspondence ([5]). To our knowledge, no R8-specific independent kinetics benchmark has been published.
Data portability is a gating item. A published Benchling connector reads XLSX exports from Biacore Insight ([6]), regulation expects validated systems and secure audit trails ([7]), and FDA guidance describes attributable, original and accurate data (ALCOA) ([8]). A feature list is not a compliance statement.
No public list price exists for any of the three systems. Public traces are service awards, grants and core-facility rates, for example $23,870 for one year of service on a Biacore SPR system at a federal lab ([9]) and $445 per day for 8K+ time at one university core ([10]). The practical recommendation is a blinded same-panel pilot with acceptance thresholds set by assay owners before data review, followed by a cost model built on accepted interactions. In an illustrative model with placeholder inputs, falling from a 95% to a 50% pass fraction roughly doubles cost per accepted interaction (about $22 to $45), more than reviewer time does.
Affinity agreement for 20 of 23 Fabs in the older Biacore 8K versus LSA study
Affinity correspondence in a 48-antibody LSA versus KinExA comparison
Pairwise interactions required for binning 100 antibodies
Introduction and Background
Antibody-discovery teams choose a binding platform for three different jobs: quantifying titer (concentration) in crude samples, confirming kinetics (association rate ka, dissociation rate kd and equilibrium constant KD) on selected leads, and sorting large panels into epitope bins. Results then have to land in an electronic lab notebook (ELN) or a registry with their context intact. The same instrument rarely wins all three, and in many labs BLI and SPR are complementary: BLI for fast titer and screening, SPR for kinetic confirmation. BLI dips biosensors into wells whereas SPR passes samples over a surface through flow channels or arrays, which is the root of most practical differences discussed below.
This report focuses on an evaluation framework: treat experimental design and data portability as acceptance gates, and do not assume that software alone substitutes for a complete instrument workflow. Where the report refers to "newer variants", it does so only to prevent mistaken procurement, and the compared models remain the Biacore 8K+, Octet R8 and LSA.
Buyer decision tree
The tree below uses three common decisions in an instrument shortlist.
- Titer or concentration in crude supernatant: start from the BLI format. Sartorius states the R8 performs quantitation of up to 96 samples in 30 minutes to 2.5 hours depending on the assay ([11]). SPR platforms can quantify too (Cytiva lists a Concentration & Potency extension), but this is not their primary design point.
- Kinetic confirmation of tens to hundreds of leads: shortlist all three, then let the pilot decide. The 8K+ and LSA are continuous-flow or flow-printed SPR; an independent four-platform study found continuous-flow SPR better at resolving slow dissociation than a fluidics-free Octet RED384 ([12]).
- Large antibody binning panel: array SPR is the design match, since full pairwise binning scales quadratically: ten antibodies need 100 interactions and 100 antibodies need 10,000 ([13]). Carterra states 384 by 384 binning in a single run ([14]); Octet Analysis Studio has a binning module ([15]); the Biacore epitope binning is an optional extension.
- Small molecules or fragments: treat the 2020 comparison's caveat seriously: small-molecule signals of only a few response units are not within the LSA specification ([16]).
Experimental design belongs inside the decision. Several authors stress that platform comparisons are meaningful only when SPR best practices are followed ([17]), so a purchase should not rest on a headline throughput number.
Scope and evidence labels
Evidence in this report is labelled by type: vendor specification or claim, vendor software documentation, independent or peer-reviewed experiment, public procurement record, and proposed pilot output. Where a vendor page is the only source, it is described as a claim.
Cytiva Biacore 8K+
Capabilities
The user manual (document 29287247, edition AC) covers the Biacore 8K and 8K+ together and describes a basic software package with optional extensions ([18]). Each of the 8 channels has two flow cells; ligand is normally attached in Fc2 while Fc1 serves as the reference ([19]). Flow settings apply to all channels in a given cycle, a constraint to remember when planning mixed-condition screens.
Cytiva's datasheet (document code CY11892-02Jun26-DF) states that the 8K+ sample hotel holds up to six trays versus two for the 8K ([20]) and quotes unattended capacity of 1536 and 4608 samples for the 8K and 8K+ ([21]). It also claims that parallel kinetics screening can identify hits from 384 samples in less than 6 hours ([22]). These are vendor claims under Cytiva's conditions.
Adoption and Lifecycle
Public procurement records show the 8K+ in use at a US federal research institute, where a service solicitation required work by a Cytiva factory-trained engineer ([23]), and a university shared resource lists 8K+ time for internal, external academic and for-profit users ([10]). An NIH shared-instrumentation award of $438,939 in 2021 funded a Biacore 8K for a university biosensor core ([24]).
Lifecycle caution: in May 2026 Cytiva announced the next generation of the Biacore 8 series ([1]). The Biacore 8S and 8S+ were described as screening 384 proteins in approximately 15 minutes, 16 times faster than the previous generation (a vendor claim) ([25]), with orders expected in summer 2026 ([26]). Cytiva's May 2026 selection guide lists 8S+ capacity of up to 12 by 384-well plates and 72 hours unattended ([27]) ([28]). Cytiva has not publicly announced the discontinuation of the 8K or 8K+, so a buyer should ask Cytiva to state the supported status, the service horizon and the software version path for the exact model quoted.
Strengths and Limitations
- Strength, data quality for small molecules: the 2020 comparison concluded that a key advantage of the 8K is its exceptionally low noise ([29]).
- Strength, extensible software: an optional API server gives automated access to run and evaluated data ([30]), and v6.0 release notes document fit mode (global, local, constant) and initial-value columns in exported result tables ([31]).
- Limitation, optional modules: epitope binning, data integration and GxP ([32]) are separate extensions, and the basic package and extensions are licensed separately through a licensing server ([33]).
- Limitation, default access model: in the basic package all folders and objects in the database are accessible to all users ([34]); restricted access for routine users and audit trails come with the GxP extension.
- Limitation, ecosystem fit: Benchling's published connector parses spreadsheet exports from Biacore Insight ([6]), but its documentation lists epitope binning and screening campaigns as not marked supported ([35]).
Sartorius Octet R8
Capabilities
Sartorius describes the Octet R8 as an eight-channel system that performs quantitation and kinetic analysis of up to 96 samples in 30 minutes to 2.5 hours, depending on the assay. Its channels can run independently for screening or in tandem, pairing the sample read with a dedicated reference ([36]). Specifications list a sample volume of 180 to 220 microlitres per well ([37]), and the biosensors are disposable and single-use, with optional reuse by regeneration.
The software story changed with Octet Software Version 14, which introduced Octet Hub as a centralized database for methods, experiments, analysis settings and results ([38]). Hub provides version history with revert and role-based access, is deployed on premises ([39]), and works exclusively with v14 applications ([40]).
Adoption and Lifecycle
A federal cancer institute's notice for an Octet R4 maintenance agreement states that Sartorius is the sole provider of service for the Octet line ([41]), which is a procurement fact worth carrying into any three-year service assumption. A university core also lists an Octet RH16 (a 16-channel sibling) at $35 per hour internal and $56 per hour external ([42]), useful as a core-facility rate pattern but not an R8 price.
Sartorius now promotes the Octet R8e as its latest BLI system ([43]), claiming reliable results with as little as 40 microlitres ([44]), 384-well plate compatibility ([45]) and 16+ hours of evaporation control ([46]). Because the R8 and R8e differ on those documented points, quotes must name the model.
Strengths and Limitations
- Strength, simple workflow: Sartorius attributes faster time to results and process economy to the fluidics-free design, a vendor attribution without an accompanying benchmark.
- Strength, binning in software: Analysis Studio annotates the binning cycle pattern and builds a traffic-light matrix and BinChart.
- Strength, fitting breadth: models include 1:1, 1:2, 2:1 and mass transport with local and global options ([47]).
- Limitation, slow dissociation and high affinity: an independent study on older BLI hardware found upward drift during dissociation, attributed mainly to evaporation ([48]), and another reported that BLI underestimated low-picomolar affinities ([4]). These findings predate the R8 and should be re-tested rather than assumed.
- Limitation, role in the workflow: the same authors conclude BLI is effective for epitope binning and qualitative affinity assessment ([49]).
- Limitation, compliance is conditional: Sartorius states that compliance depends on how the system is configured and used ([50]), and that validation activities in the v14 package are performed by the end user ([51]).
“Define thresholds before review, capture lineage into the ELN or registry, price the module list line by line, and compute cost per accepted interaction with your own pass fraction and reviewer time.
Carterra LSA
Capabilities
Carterra describes the LSA as combining flow-printing microfluidics with SPR detection. Its 96-channel printhead can be used up to four times to build a 384-ligand array ([52]); a camera images the whole chip so up to 384 real-time interactions are monitored in parallel with local referencing by neighbouring interspots ([53]). Stated workflows include capture kinetics of up to 1152 mAbs ([54]) and binning of up to 384 by 384 antibodies. Carterra lists about 200 microlitres per immobilization in printhead mode ([55]). The vendor says coupled kinetics typically needs about 200 ng of each clone and roughly 100 pmol of antigen ([56]).
Fixing an array is not free: in the 2020 study, arraying 384 ligands in four batches of 96 took approximately 2.5 hours with a 5 minute print contact time ([57]).
Adoption and Lifecycle
The product line has moved. Carterra's overview table lists the LSA beside LSAXT, Ultra and Vega ([58]). The comparison table gives the LSA analyte sensitivity as at least 1000 Da versus at least 500 Da for the LSAXT ([59]), and Carterra states the LSAXT can measure analytes down to 500 Da ([60]). The Vega (announced 11 February 2026) is described as a 48-channel platform ([2]). Some Kinetics 2.0 features are documented as exclusive to Ultra ([61]). A federal notice of intent in 2025 for an LSA XT one-year service contract used a sole-source basis with the vendor ([62]).
Strengths and Limitations
- Strength, panel scale: an industry collaboration ran all 35 anti-PD-1 antibodies on a single 384-array per chip ([63]) and resolved ten bins by classical sandwich binning ([64]).
- Strength, linked review: the binning tool links sensorgrams, heat maps, networks and a dendrogram, and the Kinetics software flags data automatically ([65]).
- Limitation, native models: Carterra's FAQ lists off-rate only, 1:1 Langmuir and mass transport models ([66]), recommends assay optimization to achieve 1:1 binding ([67]), and states that advanced models require export to third-party software ([68]).
- Limitation, slow off-rates: in a 48-antibody comparison with a solution method, 13 of 15 antibodies with solution affinity below 50 pM were at the LSA off-rate floor ([69]).
- Limitation, small analytes: the LSA's stated minimum is at least 1000 Da.
Feature Comparison
Table 1 below is a source-linked model, software and licence matrix. Each row carries the edition or retrieval anchor; "retrieved" means 2026-11-01 or 2026-11-02. Cells describe documented features, not compliance outcomes.
| Attribute | Biacore 8K+ (Cytiva) | Octet R8 (Sartorius) | Carterra LSA |
|---|---|---|---|
| Detection and fluidics | SPR, 8 channels with 2 flow cells each ([70]) | BLI, 8 channels, fluidics-free dip and read | SPR array, 96-channel printhead plus single flow cell |
| Sample format | 96 or 384-well plates; 8K+ sample hotel up to six trays | 96-well plate (to be confirmed with the vendor); R8e adds 384-well ([45]) | 96 or 384-well plates in deck bays (to be confirmed with the vendor) |
| Vendor throughput claim | 4608 samples unattended (8K+) | Up to 96 samples in 30 min to 2.5 h | 200 analytes per 24 h table value ([71]); up to 1152 captured mAbs |
| Native kinetic models in documentation | Fit mode global, local or constant exposed in result tables (v6.0 notes) ([31]) | 1:1, 1:2, 2:1, mass transport | Off-rate only, 1:1, mass transport ([66]) |
| Binning | Epitope binning item with self-self subtraction in v6.0 | Analysis Studio module | Epitope software, up to 384 by 384 ([72]) |
| Data store | SQL Server database (to be confirmed with the vendor) | Octet Hub (v14), on premises | Kinetics V.2.0.0.5078 manages .sprdata files, merges runs ([73]) |
| Export and integration | JSON/XML via optional Data Integration; application programming interface (API) server | CSV, XLSX, PDF with an API for external calls ([74]) ([75]) | Excel and SQLite export ([76]); third-party export for advanced fits |
| Regulated-use package | Optional GxP extension, validation support file | Optional GxP package with 21 CFR Part 11 software and installation, operational and performance qualification (IQ/OQ/PQ) kits | No Part 11 statement published on the Carterra pages reviewed |
| Newer family variants | Biacore 8S, 8S+ announced May 2026 ([77]) | Octet R8e | LSAXT, Ultra, Vega ([58]) |
| Public price | Quote only; service and core-rate data points only | Quote only | Quote only |
The matrix shows that "included" is the wrong default assumption for software. Cytiva lists binning, data integration and GxP as extensions, Sartorius separates Research and Compliance suites ([78]), and Carterra's public product pages do not state access-control or Part 11 capabilities. A procurement team should therefore request a line-item quote per software module and write the purchased module list into the acceptance protocol.
Review, Fitting and Data Portability
Native fitting, reference subtraction and residual review
Reference handling differs by architecture. Biacore reserves Fc1 as the reference. Octet allows flexible sample referencing and biosensor subtraction schemes ([79]). Carterra uses local referencing via interspots, and in the 2020 comparison the LSA workflow used interspot and double referencing before 1:1 fits, with spots judged by low, randomly distributed residuals ([80]). Double referencing means subtracting both a reference channel and a blank injection ([81]).
Residual review is a documented step on all three. A 2026 SPR protocol states that residuals should be randomly distributed around zero, because systematic deviations indicate a poor fit ([82]); Octet's specification lists residuals of fits among its outputs ([83]); Carterra's FAQ defines residual standard deviation, in response units, as the overall agreement between fit and data ([84]). For mass transport, the same protocol gives a screening rule that the mass transport constant should be at least 100 times kon for a 1:1 model ([85]), and it warns that very slow dissociation may be slower than an instrument can reliably detect ([86]).
Rejected-curve handling and analyst edits
Define in the protocol how rejected curves are recorded. The Biacore v6.0 notes add an Undo for removed ranges, logged in the audit trail for regulated items ([87]); the Carterra Kinetics software lets analysts add back an excluded concentration on individual sensorgrams ([88]); Carterra summary statistics use a geometric mean to average on and off-rates, which affects any attempt to reproduce an average outside the software ([89]). A floor convention also exists: antibodies at the SPR lower off-rate limit are assigned an equivalent floor value, so an exported KD for such antibodies may reflect a convention, not a measured dissociation ([90]).
Licences, access control and validation
- Cytiva: extensions are licensed separately and managed by a licensing server; the GxP extension adds restricted access for routine users and audit trails; the datasheet lists user authorization levels, audit trail, version history and electronic signatures as features ([91]).
- Sartorius: Octet Hub supports Active Directory integration ([92]); CFR software offers four user levels; data files are digitally signed and modification invalidates them ([93]); the Hub server requires Windows Server 2019, 2022 or 2025 with 4 dedicated CPU cores and 8 GB RAM minimum ([94]).
- Carterra: the public product pages describe file formats and exports, not user accounts. Treat access control as a vendor question.
The regulatory text sets the yardstick without naming any product. 21 CFR 11.10 expects validation of systems for accuracy, reliability and consistent intended performance ([7]), secure time-stamped audit trails ([95]), authority checks ([96]) and accurate, complete copies of records in human readable and electronic form ([97]). A feature list is not compliance; Sartorius itself says compliance depends on user management, e-signature and audit trail use ([98]).
Raw-to-final data lineage
FDA's data integrity guidance describes complete data as attributable, legible, contemporaneously recorded, original or a true copy, and accurate (ALCOA) ([8]), defines metadata as the contextual information required to understand data ([99]), and says a printout or static record does not preserve the dynamic format of a dynamic record ([100]). In FAIR terms, identifiers should be globally unique and persistent ([101]) and results should carry detailed provenance ([102]), although the FAIR authors say the principles are not a specification ([103]). Table 2 turns this into a lineage table to be completed per platform during the pilot.
| Lineage element | What to capture | Where to look, by documented capability |
|---|---|---|
| Sample ID and ligand ID | Registry ID, lot, plate and well, spot or sensor position | Carterra Epitope added Bay and Position information to data tables ([104]); confirm equivalent plate-map exports for the other two in the pilot |
| Surface loading | Capture or immobilization level in response units | Carterra FAQ links the interpretation of Rmax to ligand level |
| Concentration series | Analyte concentrations and excluded points | Excluded concentrations can be added back in Carterra Kinetics |
| Reference subtraction | Reference surface or sensor, blank cycles, double referencing | Documented in all three, see above ([80]) |
| Fitted model and mode | 1:1, mass transport, two-state; global, local or constant; initial values | Biacore fit mode columns exported via Excel, PDF, PPT or API; Carterra FAQ models |
| kon, koff, KD | Values with confidence or standard error, floor flags | Floor convention for slow off-rates ([90]) |
| Residuals | Residual SD or plot reference | Carterra res sd; Octet residual plots |
| Analyst edits | Who removed or restored what, with timestamps | Biacore audit-trailed undo in regulated items; Octet audit trail exportable as PDF |
| Run and version IDs | Instrument serial, software version, database or Hub project | Carterra .sprdata files and merged runs; Octet Hub v14 only |
The table is a checklist, not evidence that every cell is exportable. In particular, exporting only PDF or PowerPoint reports would not preserve reprocessable raw data, so the acceptance test should require a machine-readable export of sensorgrams plus fit parameters.
Export formats and integration
Cytiva documents Data Integration as optional export in JSON and XML and an optional API server for third parties. Benchling's connector instead consumes XLSX exports from Biacore Insight, maps them into an Allotrope Simple Model (ASM) binding affinity data model ([105]), and requires specific sheets such as Properties and Report Point Table ([106]). The open-source allotropy library converts text or Excel output into ASM JSON but does not convert proprietary or binary formats ([107]) ([108]). Allotrope's public repository makes recommended-state models public while candidate release models require membership ([109]).
Sartorius lists CSV, XLSX and PDF export with raw and processed data available for downstream analysis ([74]) and states that a software API is available, with no public specification on the product page ([75]). Carterra documents Excel and SQLite export in Kinetics 2.0 ([76]), drops support for legacy .lsadata files ([110]), and notes the software is compatible with files from all Carterra instruments. The practical rule: test the exact file that your ELN or registry will ingest, on your own data, before signing.
Performance and Benchmarks
Biacore 8K and LSA, the direct study
The 2020 SLAS Discovery study tasked two laboratories skilled in SPR with a blinded comparison ([111]); both used 23 Fabs binding immobilized CD3-epsilon ([112]). Twenty of 23 Fabs gave affinities within 1.5-fold across platforms, and the three outliers were weak-affinity Fabs ([3]). The authors attribute the mis-estimates to insufficient top analyte concentration, which was a user-influenced parameter and not an instrument limitation ([113]), and they saw no significant effect of different chips (CM5 versus HC-30M) at slow on-rates ([114]). The abstract concludes near-identical results under those conditions ([115]).
Read it with two limits. First, it tested the Biacore 8K, not the 8K+, and no BLI platform. Second, it also found complementary strengths: low noise on the 8K and one-on-many screening of a small analyte panel over hundreds of ligands on the LSA ([116]). A separate industry collaboration on anti-PD-1 antibodies found near-identical constants on flat chip types ([117]) with affinities spanning single-digit picomolar to nearly 425 nM, which challenged the methods' dynamic range ([118]).
SPR versus BLI binding kinetics
Independent BLI-versus-SPR evidence is older and hardware-specific. In a comparison on high-affinity anti-PCSK9 antibodies, continuous-flow SPR systems resolved slow dissociation better than the Octet RED384, BLI experiments were restricted to under 12 hours due to evaporation ([119]), and the authors cite lower maintenance for BLI than for fluidic SPR. Publication fees in that paper were paid by the authors' employer, a pharmaceutical company rather than an instrument vendor ([120]). A 2013 mAbs study reported that BLI consistently underestimated low-picomolar affinities, with Octet versus KinExA KD differences from 3 to 60-fold for seven antibodies ([121]), and advised optimizing the loading density of the immobilized partner. None of these studies used the R8, so the honest answer to "Octet R8 kinetics accuracy" is that the R8-specific independent benchmark was not located; the vendor offers a kinetics method development guideline referencing USP chapter 1032 ([122]).
Cross-platform checks also matter for binning. BLI response is not linearly dependent on the mass ratio, so blocking thresholds do not transfer across platforms ([123]); cross-blocking is necessary but not sufficient for two antibodies to bind the same epitope ([124]); and antibodies can compete by kinetically altering one another, so sensorgram review matters beyond a heat-map call ([125]). An older 96-antibody comparison found an SPR imaging system consumed almost 100-fold less antibody analyte than a BLI system ([126]), again on earlier hardware.
Solution-phase orthogonal checks
When an SPR array reports floor off-rates, an orthogonal method is the check. In a 48-antibody comparison of the LSA with KinExA, the affinities showed 54% correspondence ([5]) and 26 antibodies fell within 0.5 to 2-fold ([127]). KinExA throughput is much lower than SPR, and the anti-PD-1 solution mixtures were allowed to equilibrate for up to 64 hours, so orthogonal checks have a cost in time.
Data Analysis and Evidence
Public price and cost evidence
The best-supported price statement is negative: no public list price is available for the Biacore 8K+, Octet R8 or LSA. The numbers that are public are context, not instrument prices:
- Service award: a prior one-year service award of $23,870 (an aggregator-reported incumbent value, not an official award notice) on a Biacore SPR system at a federal institute, with the new requirement covering an 8K+ ([9]).
- Reseller plan: a government-schedule listing shows a FullCare service plan for the Biacore 8K+ at $47,900 each ([128]), which is a service plan, not an instrument.
- Capital grant: an NIH award of $438,939 funded a Biacore 8K core upgrade in 2021; the figure is a grant amount, not a list price.
- Core rates: $445 per day internal, $890 per day external academic and $2,150 per day for-profit for 8K+ time at one university core, with sensor chips and capture kits charged extra ([129]).
- Carterra: a one-year LSA XT service contract notice names a Summit Level plan on a sole-source basis; no amount was disclosed.
Treat all of these as inputs to ask about, not as a basis for ranking. Service costs for Octet are subject to the sole-provider statement.
Cost per accepted interaction
The cost-per-accepted-interaction calculation is:
- Accepted interactions = attempted interactions x observed pass fraction.
- Fully loaded cost per accepted interaction = (annualized instrument + service + software + consumables + preparation + review + reruns + integration) / accepted interactions.
As a hypothetical example, a lab attempts 20,000 interactions per year. Placeholder inputs: $300,000 per year for annualized instrument, service, software and integration combined; $4 per attempt for consumables and preparation; 1.5 minutes of reviewer time per attempted interaction at $90 per hour; and half of rejected interactions rerun at the $4 attempt cost. None of these figures is a vendor price or a platform benchmark. Table 3 shows the sensitivity of cost per accepted interaction to pass fraction and reviewer time.
| Pass fraction | Accepted interactions per year | 0.5 min review per attempt | 1.5 min review per attempt | 3.0 min review per attempt |
|---|---|---|---|---|
| 95% | 19,000 | $20.89 | $22.47 | $24.84 |
| 85% | 17,000 | $23.59 | $25.35 | $28.00 |
| 70% | 14,000 | $29.07 | $31.21 | $34.43 |
| 50% | 10,000 | $41.50 | $44.50 | $49.00 |
In this illustration the pass fraction dominates: moving from 95% to 50% raises cost per accepted interaction by about 98% at 1.5 minutes of review, whereas moving review time from 0.5 to 3.0 minutes at 85% passes raises it by about 19%. Fixed costs are about $17.65 of the $25.35 at 85% passes. Channel count does not appear anywhere in the formula, which is the point: useful throughput is accepted interactions, and a platform with more channels but a lower pass fraction on your molecules can cost more per accepted result.
Same-panel pilot with predefined acceptance criteria
Use one panel on all three candidate systems, run by operators who know the platform, with blinded review. The assay owners, not the vendors, should set thresholds before any data are opened. The form below carries no invented benchmark values.
- Panel: include weak, mid and tight binders, a negative control set, replicate pairs, and at least some antibodies with known solution-phase KD.
- Design freedom: document ligand level, analyte range and dissociation time per platform, since the 2020 study showed user choices drive outcomes.
- Precision threshold: assay owner sets maximum fold or coefficient of variation between replicates, written before review.
- Concordance threshold: assay owner sets acceptable fold difference versus the reference method per affinity band, including a rule for floor off-rates.
- Blinded review: reviewers see fits without platform labels; record accept, reject and reason.
- Rejection log: count rejects by cause (poor fit, mass transport, surface loss, carryover).
- Reviewer time: log minutes per interaction to feed the cost model.
- Portability test: export raw sensorgrams, fit parameters and identifiers into the target ELN or registry and reconstruct three results independently.
- Version capture: record software and firmware versions on every run.
Include weak, mid and tight binders, negative controls, replicate pairs and antibodies with known solution-phase KD.
Document ligand level, analyte range and dissociation time for each platform.
Assay owners set precision and concordance thresholds before any data are opened.
Review fits without platform labels and record accept, reject and reason.
Count rejection causes and log minutes per interaction to feed the cost model.
Export raw sensorgrams, fit parameters and identifiers to the target ELN or registry, then independently reconstruct three results.
Implications and Future Directions
Three shifts matter for the next procurement cycle. First, vendors have recently moved their families forward: the Biacore 8S and 8S+ with Insight v7.0 and Sensorgram comparison, which converts interaction profiles into a similarity score ([130]); the Octet R8e; and the Carterra Vega with a 48-channel format and a claim of more than 20,000 small-molecule interactions per day ([131]). Buyers who specify a model without a configuration may receive an unexpected variant, and an evaluation done on an older model may not transfer.
Second, data portability has become an architectural question. The ASM route requires text or Excel exports, so file-based export from an instrument is a gating item, and the Benchling documentation shows that connector coverage differs by analysis type. Informatics teams should record which ASM release and which export template were used.
Third, regulated use is a configuration and validation exercise. Octet validation activities are performed by the end user, Cytiva offers a validation support file ([132]), and the regulation expects validated systems, authority checks and audit trails. This is work for assay owners, quality and informatics together, not a feature to be assumed.
Evidence gaps and vendor questions
Known gaps: no independent head-to-head including the Octet R8, the 8K+ and the LSA; no R8-specific independent kinetics benchmark; no public price for any of the three; no Part 11 or access-control statement on Carterra's public product pages; no public Octet export schema in the user guide. Ask each vendor:
- Edition: which software, firmware and database versions are in the quote, and what is the supported lifecycle for the quoted model after the newer family variants?
- Modules: a line-item list of licences (binning, integration, GxP) with seat or server terms.
- Export schema: a sample export of raw sensorgrams, fit parameters, model metadata and sample identifiers, with schema documentation.
- API versus file: is the integration a documented API or file export, and what are the version dependencies?
- Audit and access: how are analyst edits and rejected curves recorded, and can the record be exported?
- Consumption: assay volume and consumable use for the pilot design.
- Validation: what the vendor provides for installation and performance qualification, and what the buyer must execute.
Frequently Asked Questions (FAQs)
Biacore 8K+ vs Carterra LSA: which gives closer kinetics?
The direct 2020 study of the Biacore 8K and LSA found 20 of 23 Fab affinities within 1.5-fold; it was conducted with skilled users, and does not extend automatically to the 8K+ or to other analytes.
Octet R8 vs Biacore 8K: is BLI as accurate as SPR?
Independent evidence on older Octet hardware shows weaker resolution of slow dissociation than continuous-flow SPR and underestimation of low-picomolar affinities. No R8-specific independent data have been published to our knowledge.
Carterra LSA vs Octet R8 throughput: which is faster?
They are not directly comparable. The R8 claim is up to 96 samples in 30 minutes to 2.5 hours; the LSA claim is screening up to 1152 mAbs. Compare accepted interactions per week on your panel.
What are the main high-throughput SPR instruments for comparison?
In this report's scope, the Biacore 8 series (8K, 8K+, and the newer 8S and 8S+), the Carterra LSA family (LSA, LSAXT, Ultra and Vega) and, as a BLI alternative, the Octet R series (R8 and R8e). Models differ in ligand capacity, sample handling and software, so compare them on your panel, not on the number of channels.
What is the best SPR instrument for antibody screening?
For very large panels, array SPR (LSA family) is designed for it; for flexibility including small molecules, flow-based SPR has documented low noise. Both need a pilot.
How much does a Biacore 8K+ or Carterra LSA cost?
No public list price is available for either. Public data points are service awards, reseller service plans and core rates, none of which is an instrument price.
Are the systems 21 CFR Part 11 compliant?
A feature set is not a compliance statement. Cytiva describes an optional GxP extension; Sartorius offers a GxP package and says compliance depends on configuration.
Conclusion
The three platforms answer different questions. The Biacore 8K+ pairs eight-channel flow SPR with optional extensions for binning, integration and regulated use. The Octet R8 offers fast, dip-and-read BLI with software that moved to a governed Hub in version 14. The Carterra LSA offers array-scale SPR for kinetics and binning with a smaller set of native models and an export route for the rest. Independent evidence is thinner than the marketing: the one direct SPR study compares two platforms using one older Biacore model, and BLI benchmarks predate the R8.
A defensible purchase therefore rests on a pilot, not a table. Define thresholds before review, capture lineage into the ELN or registry, price the module list line by line, and compute cost per accepted interaction with your own pass fraction and reviewer time. Re-check model names, software editions and lifecycle statements at quote time, since all three vendors have newer family members.
About IntuitionLabs
Build practical AI for pharma and biotech with IntuitionLabs. We help life-science teams turn complex information and workflows into useful software, governed knowledge systems and AI tools.
IntuitionLabs is an AI consulting, custom software development and data engineering firm serving pharmaceutical, biotechnology, medical-device and other life-science organizations. We work with clinical, regulatory, medical-affairs, commercial, quality and IT teams to connect technology decisions with the work people need to accomplish.
AI consulting and adoption
Our AI enablement services cover readiness assessments, use-case selection, governance and policies, team workshops, adoption measurement and ongoing advisory support. We help organizations structure the information layer behind AI: source material, context, permissions and maintained knowledge that make generated answers useful and reviewable. Private LLM inference and hosted AI options support teams evaluating how to operate AI with appropriate control over their data and infrastructure.
Software, data and life-science workflows
IntuitionLabs develops custom software for pharma and biotech, integrates enterprise systems, and builds data engineering and business intelligence solutions. Areas of focus include AI agents, regulatory research, medical writing, medical affairs, CMC information, competitive intelligence and clinical-document workflows. Our eTMF intelligence work includes cross-system reconciliation and inspection-readiness support.
Enterprise platforms and regulated delivery
We provide Veeva services, application support, managed services, integrations and custom applications, alongside enterprise content work involving platforms such as Egnyte. For regulated workflows, our services include GxP enablement, computer-system validation and software development addressing 21 CFR Part 11 requirements. The applicable controls, validation responsibilities and acceptance criteria are defined for each engagement.
Work with IntuitionLabs
Explore AI enablement, pharma and biotech software development, data engineering and BI, and Veeva services. Contact IntuitionLabs to discuss your workflow, information sources and implementation needs.
IntuitionLabs publishes educational research to help life-science teams make informed technology decisions. Coverage of a product or organization does not imply a client relationship, endorsement or partnership.
Sources / 132
Need Expert Guidance on This Topic?
Let's discuss how IntuitionLabs can help you navigate the challenges covered in this article.
I'm Adrien Laurent, Founder & CEO of IntuitionLabs. With 25+ years of experience in enterprise software development, I specialize in creating custom AI solutions for the pharmaceutical and life science industries.
The information contained in this document is provided for educational and informational purposes only. We make no representations or warranties of any kind, express or implied, about the completeness, accuracy, reliability, suitability, or availability of the information contained herein. Any reliance you place on such information is strictly at your own risk. In no event will IntuitionLabs.ai or its representatives be liable for any loss or damage including without limitation, indirect or consequential loss or damage, or any loss or damage whatsoever arising from the use of information presented in this document. This document may contain content generated with the assistance of artificial intelligence technologies. AI-generated content may contain errors, omissions, or inaccuracies. Readers are advised to independently verify any critical information before acting upon it. All product names, logos, brands, trademarks, and registered trademarks mentioned in this document are the property of their respective owners. All company, product, and service names used in this document are for identification purposes only. Use of these names, logos, trademarks, and brands does not imply endorsement by the respective trademark holders. IntuitionLabs.ai is an AI software development company specializing in helping life-science companies implement and leverage artificial intelligence solutions. Founded in 2023 by Adrien Laurent and based in San Jose, California. This document does not constitute professional or legal advice. For specific guidance related to your business needs, please consult with appropriate qualified professionals.