← back to ranking

TXG · 10x Genomics, Inc.

Medical - Healthcare Information Services · mkt cap $4.4B · calls: Q1 FY2026 vs Q4 FY2025
41.0 conviction · conf-adj 41

conf 2/10 partial

enthusiasm:27.0 · trend:8 · quantifies:0 · impact:0 · under_radar:5 · credibility:0 · business_impact:4 · disruption:-6 · commitment:0 · confirmation:3

Enthusiasm latest 9 / prev 8 (rising)

10x Genomics frames itself as the essential data-infrastructure layer for AI-driven biology, arguing that scaling laws governing LLM training apply equally to biological datasets and that demand for its single-cell and spatial products will grow exponentially as AI model development scales — with Atara explicitly launched as a billion-cell-per-year AI training engine and a new addressable market in its own right. Enthusiasm escalated meaningfully from Q4 2025 to Q1 2026, driven by three new high-profile AI partnerships in a single quarter and management asserting that virtually every large customer project now has AI as its primary driver. Credibility is constrained by the self-admitted 'relatively small percentage' AI revenue base in 2025, the complete absence of any AI-specific revenue, booking, or unit guidance across both calls, and the heavily theoretical framing that underpins the entire thesis.

GROUNDED NEXT-FY IMPACT vs CONSENSUS

Grounded on actual base — revenue $0.6B · net income $-0.0B · net margin -6.8% · diluted EPS -0.35

These are next-fiscal-year annual uplift estimates, not next-quarter numbers.

Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: % · vs analysts: unclear · priced in: medium (model's call-read: low; verdict above is the hard-data one used for ranking) · confidence: 2/10

ClaimFigureArithmeticNext-FY Rev %Next-FY EPS %
CZI Biohub $100M virtual-biology initiative
other · soft		$100M (partner budget, not TXG revenue)$100M is CZI/Biohub's OWN initiative budget, not TXG revenue/bookings. No take-rate or TXG capture share disclosed. Illustratively 100*100M/642.823M=15.6% of FY25 revenue IF fully captured, but it is a third-party commitment with no attribution path -> unattributable. EPS null (loss-maker).
Biooptimus Stella up to 100,000 patient tissue specimens
engagement · soft		up to 100,000 specimens (program scope)Customer program scope, not TXG revenue. No price-per-specimen, consumable pull-through, or TXG share disclosed -> no revenue figure derivable. EPS null (loss-maker).
Atara throughput (800 whole-transcriptome samples/yr/instrument; billion-cell/yr)
other · soft		800 samples/yr/instrument; billion-cell/yrThroughput/capacity SPEC cited to attract AI customers, no revenue attached. No installed-instrument count, ASP, or attach rate disclosed -> cannot convert capability to dollars.
AI-driven revenue from relatively small 2025 % base, growing
revenue · soft		relatively small percentage (no % or $ disclosed)Management cites AI-driven revenue as a small share of total with growth expected, but gives no percentage or dollar base -> cannot compute 100*delta$/642.823M or EPS bridge. soft=true (unanchored %).
Forward qualitative AI demand (exponential scaling, no ceiling)
engagement · soft		no $/% quantifiedStatements on exponential AI-driven research demand, scaling laws, and Atara as discovery engine contain no attributable TXG revenue, bookings, or cost savings -> null; soft=true.

Assumptions: No next-FY TXG revenue or cost $ attributable from disclosed claims. Tax 21% and current net margin (-6.77%) defaults NOT applied because no claim carries a TXG dollar figure or convertible base. All 'quantified' items are explicitly third-party/customer/product-spec numbers with no disclosed take-rate, ASP, or attach rate. Bookings!=revenue: CZI $100M treated as partner budget, zero TXG recognition. EPS sizing vs consensus adjusted EPS not attempted: FY25 net income=-$43.544M and FY26 consensus losses widen -> EPS-uplift % meaningless. Side=adopter: TXG sells biology data-generation tools (instruments/consumables) whose demand AI lifts; it does NOT sell AI compute/chips/infrastructure into the buildout, so no supplier-side revenue.

Top line: No quantified TXG topline impact can be sized. All three quantified figures ($100M CZI budget, 100k Stella specimens, Atara's billion-cell/800-sample throughput) are partner-budget, customer-program-scope, or product-capability numbers with NO TXG revenue, take-rate, or ASP attached. CZI $100M would be 15.6% of revenue even if 100% flowed to TXG, but it is a third party's spend with no disclosed capture path. Forward narrative ('exponential demand', 'no credible ceiling', AI revenue a 'relatively small percentage' base) is directionally positive but entirely unanchored. Aggregate est_rev_uplift_pct=null; supplier-side also null.

Bottom line: No quantified opex/productivity savings. TXG is loss-making (net income -$43.544M, net margin -6.77% on $642.823M revenue; gross profit $443.881M, operating income -$110.885M), so any EPS-uplift % off this base is denominator-artifact/meaningless per guardrail -> est_eps_uplift_pct=null. No anchored after-tax saving or incremental net income exists to flow through. AI is described as additive demand for data-generation consumables, not a margin/cost program.

[impact n/m (all claims soft/unanchored); EPS uplift n/m (loss-making base)] Sized AI uplift is null (0% hard add) because management disclosed no TXG AI revenue figure. Consensus models FY26 revenue at $612.9M — a 4.7% DECLINE from the $642.8M base — then $664.5M (+8.4%) in FY27, with losses widening (EPS -0.43 -> -0.76 -> -0.64; net income -$49.4M -> -$104.3M -> -$90.4M). So consensus bakes in essentially zero AI premium. Management simultaneously describes 'exponential' AI-driven data demand with 'no credible ceiling'. If that demand even partially materializes it is clearly NOT in a declining-revenue consensus -> priced_in=low. This is a directional read only: TXG disclosed no AI revenue figure, so the size of the gap cannot be computed and conviction is correspondingly weak.

MODEL CONSENSUS (impact)

partial

Both null all figures (no anchored TXG $); merged X's fuller 5-entry soft list with Y's sounder side/priced_in verdicts.

Conflicts reconciled
FieldOpus 4.8GPT-5.5
Rev uplift %––
EPS uplift %––
Priced inlow–
vs analystsunclear–
Confidence2–
Top lineNo quantified TXG topline impact can be sized. All three figures ($100M CZI budget, 100k Stella specimens, Atara's billion-cell/800-sample throughput) are explicitly partner-budget, customer-program-scope, or product-capability numbers with NO TXG revenue, take-rate, or ASP attached. The CZI $100M would be only 15.6% of revenue even if 100% flowed to TXG, but it is a third party's spend with no disclosed capture path. The forward narrative ('exponential increase in demand', 'no credible ceiling', AI revenue a 'relatively small percentage' base in 2025) is directionally positive but entirely unanchored — management gave no AI revenue number, run-rate, or growth figure to model.–
Bottom lineEPS uplift is undefined: TXG is loss-making (net income -$43.544M, net margin -6.77%), so any EPS-uplift % off this base is meaningless per the guardrail (est_eps_uplift_pct = null). Even setting that aside, no anchored after-tax saving or incremental net income exists to flow through, since no claim carries a dollar figure. AI is described as additive demand for data-generation consumables, not a margin/cost program.–
ReasoningConsensus models FY2026 revenue at $612.9M — a 4.7% DECLINE from the $642.8M base — then $664.5M (+8.4%) in FY2027, with losses WIDENING (EPS -0.43 -> -0.76 -> -0.64). So consensus bakes in essentially zero AI premium near-term. Management simultaneously describes 'exponential' AI-driven data demand with 'no cap'/'no credible ceiling'. If that demand even partially materializes it is clearly NOT in a declining-revenue consensus, hence priced_in = low. BUT this is a directional read only — TXG disclosed no AI revenue figure, so the size of the gap cannot be computed and conviction is correspondingly weak.–

Rows highlighted where the two models disagreed.

QUANTIFICATIONS
CZI Biohub virtual biology initiative budget (partner commitment, not TXG revenue): $100 million (announced Q1 FY2026, topline (indirect demand signal only; no TXG revenue figure given))
“And just last week, the Chan Zuckerberg Biohub announced its new $100 million virtual biology initiative. The goal of the effort is to build AI models that can accurately simulate cells and tissues in silico.”
Biooptimus Stella patient tissue specimens (customer program scope, not TXG revenue): up to 100,000 patient tissue specimens across three continents (forward-looking, announced Q1 FY2026, topline (indirect demand signal only; no TXG revenue figure given))
“Biooptimus is building Stella, a global initiative aiming to profile up to 100 thousand patient tissue specimens across three continents. The goal is to build a data backbone for a world model of biology.”
Atara throughput capacity positioned as AI training data engine (product spec, not revenue): up to 800 whole-transcriptome samples per year per instrument; 'billion-cell-per-year capability' (at launch, Q1 FY2026, topline (capability spec cited to attract AI customers; no revenue attached))
“large AI models represent an entirely new opportunity, where Atara's billion-cell-per-year capability enables creation of virtual models at a new level of scale and sophistication.”
PAST (realized)
CURRENT (now)
FORWARD (guidance)
TRACK RECORD — PROMISE vs DELIVERY

—/100 (no quantified promises)   no-quantified-promises  6 calls reviewed

Over six calls management repeatedly frames AI/virtual-cell tailwinds and cites partner programs (CZI Billion/Building Cell, Arc, Biooptimus) plus ML-enabled products, but never sets its own numeric AI targets with deadlines (no dated cell/model/revenue/productivity commitments). Progress is narrated qualitatively or via third-party metrics, so there is no TXG quantified AI promise track record to score.

PRICED-IN (REFINED)
MEDIUM

Est. revisions flat  ·  Fwd P/E -81.8  ·  EV/Sales 6.3x

AI claim maps to Consumables, Service, Product And Service Revenue

Analyst sentiment is only mildly improving (buys 4→5, holds 12→10) while recent price targets fell (lastMonthAvg 22 vs lastQuarterAvg 26) and forward revenue estimates slip in 2026 before a modest 2027 rebound—consensus is not aggressively baking in AI-driven growth. However, at ~6.3x EV/Sales on ~flat TTM revenue and persistent losses, valuation is not cheap for a mature-profile med-tech multiple, and the stock at ~$35 trades well above street targets (~$20–26), implying market optimism beyond published estimates. AI upside would most plausibly flow through Consumables and Service (recurring pull-through and analysis/software), but flat revisions plus a already-stretched sales multiple yield a mixed priced-in read: not fully reflected in estimates, yet partially reflected in the premium valuation.
COVERAGE — ENTHUSIASM TRAJECTORY + CATALYSTS
4Q4 FY20247Q1 FY20258Q2 FY20258Q3 FY20259Q4 FY20259Q1 FY2026

AI enthusiasm across 6 calls — trend ↗ rising

From product-embedded ML to AI as data-demand driver, named virtual-cell partnerships, Claude analysis, and Atara as structural tailwind.

RECENT AI CATALYSTS & NEWS
BUSINESS IMPACT - QUALITATIVE MATERIALITY

5/10 qualitative impact   moderate  medium-term · soft evidence

Where AI matters: consumables/instruments pull-through from customer AI biology datasets

AI is framed as a structural demand driver for single-cell/spatial data and new products (Flex Apex, Atara), but FY25 AI-linked revenue was a small, undisclosed share with no bookings/revenue guidance and only indirect partner metrics—credible strategic fit, not yet a proven P&L lever for TXG's own business.

Caveats: Upside is mostly customer AI capex/omics demand (supplier-side), not documented internal AI productivity or margin gains; No quantified AI revenue, TAM, or FY26 uplift despite rising rhetoric; In silico perturbation/virtual cells could eventually compress wet-lab units if models prove predictive enough; Consensus implies flat-to-down FY26 revenue despite AI narrative—execution risk separate from AI thesis

AI DISRUPTION / CANNIBALIZATION RISK  tailwind · 4/10

Long-run risk that mature in silico/virtual-cell models curb large perturbation-screening volumes, but management and the scaling-laws thesis argue models still require massive wet-lab training/validation data; near-to-medium term AI expands experiment scale more than it commoditizes TXG's core assay and consumables model.

OPTIONS / MARKET STRUCTURE

option liquidity: fair

ATM IV
—
TYPICAL BID-ASK
—
OPEN INTEREST
—

proxy inputs — dollar-ADV $89M · beta 2.051 · px $34.75

source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.

CONFIRMATION — INSIDERS · 13F · LANGUAGE
Mixed — insiders selling, institutions adding, management language 5/10 measured.
INSIDERS selling 13 open-market sell(s) vs 0 buy(s) — net distribution
INSTITUTIONS (13F) adding as of 2026-03-31: 57 new / 48 closed positions; 174 increased / 103 reduced; institutional ownership +0.22pp; +9 net 13F holders
MGMT LANGUAGE 5/10 measured AI as data tailwind: some present-tense demand signals, but belief- and opportunity-heavy; no AI revenue metrics.
commit “Products like FLEX Apex and now Atara exemplify this imperative and are seeing intense interest from customers building AI models of biology.”
commit “As AI models improve, we expect an exponential increase in the demand for the kind of data our technologies produce.”
commit “From the beginning, we have built our platforms precisely for that purpose.”
VERBATIM AI QUOTES
“It is poised to become the large-scale discovery engine—whether for basic research, AI-driven science, or translational applications.”
— Serge Saxonov, Q1 FY2026
“And, critically, large AI models represent an entirely new opportunity, where Atara's billion-cell-per-year capability enables creation of virtual models at a new level of scale and sophistication.”
— Serge Saxonov, Q1 FY2026
“As we have been saying for some time, we believe AI represents a significant and structural tailwind for our business. The potential to transform our understanding of biology is enormous, but progress depends critically on generating vastly more of the right kinds of data. From the beginning, we have built our platforms precisely for that purpose.”
— Serge Saxonov, Q1 FY2026
“Products like FLEX Apex and now Atara exemplify this imperative and are seeing intense interest from customers building AI models of biology.”
— Serge Saxonov, Q1 FY2026
“There is increasing evidence that scaling laws apply in biology just like they do in other domains. What we now need is many orders of magnitude more data—specifically of molecules, cells, and tissues across vast numbers of contexts.”
— Serge Saxonov, Q1 FY2026
“As AI models improve, we expect an exponential increase in the demand for the kind of data our technologies produce. This reinforces our conviction in the importance of what we are building and the vast size of the opportunity ahead.”
— Serge Saxonov, Q1 FY2026
“AI is poised to fundamentally reshape how science is done, and large-scale, high-quality biological data sits at the center of that transformation.”
— Serge Saxonov, Q1 FY2026
“And just last week, the Chan Zuckerberg Biohub announced its new $100 million virtual biology initiative. The goal of the effort is to build AI models that can accurately simulate cells and tissues in silico.”
— Serge Saxonov, Q1 FY2026
“This initiative is galvanizing the broader scientific ecosystem with an aligned view that this direction is the next grand challenge for advancing science and medicine, analogous to the Human Genome Project and its role in catalyzing the genomics era.”
— Serge Saxonov, Q1 FY2026
“We fundamentally believe and are seeing that AI is a structural tailwind to our business.”
— Serge Saxonov, Q1 FY2026
“I do not think at this stage there is a large project where AI is not a big driver, if not the biggest driver. Even smaller-scale experiments are often performed with an eye toward feeding data into AI models and scaling up down the road.”
— Serge Saxonov, Q1 FY2026
“So when we look at our business, this AI wave is lifting all of it across products and applications, and in many ways, that is by design.”
— Serge Saxonov, Q1 FY2026
“Through the market lens, we expect an exponential increase in AI-driven research and scaling of translational cohorts to really take off as we look to next year.”
— Serge Saxonov, Q1 FY2026
“there has been rapid parallel progress in AI and in the technologies used to measure biology. These two trends are highly complementary. Advances in single cell and spatial technologies have increased scale, lowered costs and made it possible to generate very large, high-quality biological data sets, while advances in AI are creating new demand for that data.”
— Serge Saxonov, Q4 FY2025
“Importantly, this represents a shift in how research is conducted with AI increasingly acting as a driver of data generation rather than just a downstream analysis tool.”
— Serge Saxonov, Q4 FY2025
“We're supporting the Chan Zuckerberg Initiative's Billion Cells project, which is generating unprecedented volumes of single cell data to fuel AI-driven biological discovery. We're also working with the Arc Institute on the Virtual Cell Atlas using large-scale perturbation data generated on our platforms to train and validate next-generation models of cell behavior.”
— Serge Saxonov, Q4 FY2025
“Over a short time, we believe Flex has become a foundational assay for several of the most important growth areas in the field, including large-scale AI and virtual cell efforts, translational cohort studies and biopharma discovery and development workflows.”
— Serge Saxonov, Q4 FY2025
“there's a big wave of AI-driven projects for [ Perturb-seq ] type applications. And our products, especially Flex, are incredibly well suited for that purpose. And you can actually see that now coming out in preprint on the publications validating the premise.”
— Serge Saxonov, Q4 FY2025
“They were meaningful, a relatively small percentage of our business last year, and we expect it to keep growing going forward. And I think what's particularly exciting to us is that there is — as you look to the future, the upside is enormous. There's like really no credible ceiling to how much data people are looking to generate and how much data would be useful to generate for these AI models.”
— Serge Saxonov, Q4 FY2025
“I think we're just at the very early stages of the sort of the scaling revolution for generating AI. It's sort of — it's very analogous to what has happened in other domains where AI has been applied, and there is good reasons to think why it might be even more relevant and powerful for these kinds of biological data sets where the complexity is just enormous.”
— Serge Saxonov, Q4 FY2025
“a lot of it is actually being driven by Flex Apex. It is the perfect assay for all these projects for driving perturbation screening for doing it across many different cell types and tissues. But it's still very, very early days. And like I said earlier, what's particularly exciting here is that as we look to the future, there is really no cap to the upside here.”
— Serge Saxonov, Q4 FY2025
“on the question of sort of AI customers and those applications, the — like predominantly, the right solution there is Flex Apex that's resonating really, really well for a number of reasons. It's incredibly scalable. It has huge sensitivity — like really, really good sensitivity, it's really robust, works across many different cell types and tissues.”
— Serge Saxonov, Q4 FY2025
ANALYST QUESTIONS ON AI
Q (Q1 FY2026, Matthew Larew (William Blair)): Serge, you called out one of the new TAMs around AI. I think last quarter you mentioned that was a relatively low percentage of revenue, but you referenced today a number of new initiatives you are involved in, both on the single cell and spatial side. As you have had those discussions and gotten some sense for how customers are going to be building out their plans, what is your sense for what ultimately that TAM could look like and how it might grow over time?
A: Saxonov declined to size the TAM numerically, instead arguing AI is now so pervasive that 'I do not think at this stage there is a large project where AI is not a big driver, if not the biggest driver' and that 'this AI wave is lifting all of it across products and applications.' He framed the inability to isolate AI revenue as a feature — the whole business is benefiting — and reiterated structural tailwind language.
Q (Q4 FY2025, Tycho Peterson (Jefferies)): give us a quick walk on what you're baking in for academic and pharma, in particular, around some of these larger [ Perturb ] type studies
A: Saxonov confirmed 'a big wave of AI-driven projects for [Perturb-seq] type applications,' called Flex 'incredibly well suited for that purpose,' and acknowledged AI-driven work was 'meaningful, a relatively small percentage of our business last year.' He emphasized there is 'really no credible ceiling to how much data people are looking to generate' but gave no revenue or booking figure.
Q (Q4 FY2025, Matthew Larew (William Blair)): One is for Adam, which is if you could quantify at all either revenue orders related to AI in '25 or the expectation in '26. The second part, Serge, is for you, which is, I guess, a bit higher level. It sounds like demand right now is in service of larger projects, virtual soft and nation models. I guess I'm curious if you expect that work that demand to be iterative over time where customers are constantly building models based in part on large perturbation sets rather than sort of a one and done. That's kind of the first part. And last year, there were some papers out around the idea of in silico perturbation. So I'm curious, how you see that complementing or competing? And then the third, I guess, higher level on Serge is you have a network of CRO partners around the world and certainly, 10x can -- you're an expert user of your own products. So just as you have some of these more nontraditional companies or groups of people entering the space and building models, maybe if the customer group might shift at all for your products and maybe if that alleviates the capital constraints that some smaller customers might face if it becomes more outsourced.
A: Adam Taich did not quantify AI revenue. Saxonov refused to size it, said AI demand is 'continuous' and 'iterative' (not one-and-done), dismissed in silico perturbation as a near-term threat ('biology is just like increasingly complex, and we're very, very, very far from understanding it'), and confirmed exploring service offerings for large-dataset customers. Closed with 'it's still very, very early days' and 'there is really no cap to the upside here.'
Q (Q4 FY2025, Salem Salem (Barclays)): Are you able to kind of use the proposition of this new sort of integration to win over these new AI customers now with kind of the promise of providing something even higher throughput down the line?
A: Saxonov said the scale-technology integration is not yet material to revenue and the 'predominantly the right solution' for AI customers today is Flex Apex, pointing to publications validating its fit for perturbation-scale workloads.