Paige AI pricing

About

Paige is a digital-pathology company that builds FDA-cleared and CE-marked AI for cancer diagnosis. Spun out of Memorial Sloan Kettering Cancer Center in 2017 (co-founded by Thomas Fuchs), it became the first company to receive FDA authorization for an AI product in pathology — Paige Prostate Detect, cleared via the De Novo pathway in September 2021 to aid the primary diagnosis of prostate cancer. The clinical portfolio now spans the Paige Prostate, Breast, GI, and PanCancer suites, delivered through the FullFocus® FDA-cleared whole-slide-image viewer and FullFolio image-management system, with 13 CE-IVDD and 13 UKCA marks and three FDA Breakthrough Designations.

A second business serves pharma and life sciences: OmniScreen™ (1,600+ biomarkers detected from H&E alone), the Virchow and PRISM foundation models (co-developed with Microsoft Research, trained on ~1.5M+ MSK slides), and drug-development/trial-optimization services. Paige raised more than $220M (a $125M+ Series C led by KKR with Casdin Capital and Johnson & Johnson Innovation). In August 2025, Tempus AI acquired Paige for $81.25M in stock — a steep discount to that funding — primarily for Paige’s nearly 7 million digitized pathology slides and its foundation-model team.

For the most current information, visit Paige. Note: there is no public pricing page — the /pricing URL returns a 404.

Pricing summary : How Paige AI’s pricing model works

Paige is fully sales-led — there is no published rate card, no self-serve signup, and the /pricing URL 404s. Every clinical and pharma surface routes to “Request a Trial” or “Contact Us.” Pricing is quoted per deal, which is normal for regulated diagnostics that require a laboratory-information-system (LIS) integration.

What pricing is knowable comes from the UK’s NICE Medtech innovation briefing (MIB280), which documents Paige Prostate’s commercial model. Clinical AI is sold as a per-slide SaaS subscription that “typically starts at £1 per slide” and rises with the lab’s biopsy volume, the number of biopsies per slide, slides per case, and cloud-storage usage. On top of the subscription sits a one-time LIS-integration fee that “typically starts at £15,000,” varying by integration complexity and LIS provider. Initial training is bundled into the subscription; cloud storage/archiving is an optional add-on; and NICE notes Paige is developing a per-case pricing model. The pharma/life-sciences side (OmniScreen, Virchow/PRISM licensing, drug-development services) is custom-contracted, not priced per slide.

What makes this different: unlike the token- and request-metered AI tools elsewhere in this corpus, Paige’s value metric is the slide (and, increasingly, the case) — the unit a pathology lab already counts. Because the product carries the first FDA authorization in AI pathology, the meter isn’t a generic API call but a regulated diagnostic act: each billed slide is a clinical read, which is why the model couples a usage meter to an enterprise integration fee and an annual commitment rather than a credit card.

Pricing by product

Sales motions across products: sales-led only — “Request a Trial”/“Contact Us” → scoping → LIS integration → annual subscription on the clinical side; bespoke licensing/services contracts on the pharma side. No free tier, no self-serve, no monthly card checkout. Figures are NICE-reported (UK) reference points, not an official Paige rate card.

Hidden costs : What Paige AI users actually pay

For a lab, the real cost is shaped less by the headline ~£1/slide than by the volume multipliers and the integration project. NICE is explicit that the per-slide price increases with biopsy volume, biopsies per slide, and slides per case — so a complex case with several slides costs a multiple of the sticker. On top sits the one-time LIS-integration fee from ~£15,000, and cloud storage/archiving is a separate optional line.

Other things to budget for on the clinical side: deployments are annual SaaS contracts via direct sales, not month-to-month pilots, and Paige routes much of its clinical delivery through digital-pathology partners (Sectra, Roche, Indica Labs, PathPresenter, and others), whose scanner/viewer and storage costs can sit alongside the Paige subscription. On the pharma side, every figure is bespoke — foundation-model licensing and biomarker-discovery services are scoped per engagement, so there is no published reference at all.

Want to estimate your own Paige AI bill? Use the Paige AI pricing calculator to model your costs based on slide and case volume.

Pricing evolution : Paige AI pricing history and changes

Cadence

Tracked range: 2021–present. Paige has never published a public price list, so there are no Wayback price snapshots to chart — the per-slide and integration figures come from the UK’s NICE Medtech briefing, and the evolution here is regulatory and corporate, not posted rate cards.

Notable changes

• 2021-09 — Paige Prostate Detect wins FDA De Novo authorization, the first AI product cleared in pathology — the regulatory event that underpins the sales-led clinical model.
• 2024 — NICE MIB280 documents the commercial model: per-slide SaaS subscription from ~£1/slide, one-time LIS integration from ~£15,000, bundled training, optional cloud storage, and a per-case model in development.
• 2025-08 — Tempus AI acquires Paige for $81.25M in stock (plus assumption of Azure commitments), a steep markdown to Paige’s $220M+ raised; the strategic value is the ~7M-slide dataset and foundation models.

What’s unique : Paige AI’s distinctive pricing mechanics

1. The meter is a regulated diagnostic act. Paige bills per slide (and is moving toward per case) — but each billed slide is a clinical read backed by the first FDA authorization in AI pathology. That makes the usage unit a regulated diagnostic event, not a generic API call, which is why it travels with an integration fee and annual contract rather than self-serve checkout.

2. Usage meter bolted to an enterprise integration fee. The ~£1/slide subscription only works once Paige is wired into the lab’s LIS, so the model deliberately pairs a low-looking per-slide rate with a one-time ~£15,000 integration project. The slide rate is the recurring meter; the integration fee is the gate.

3. Two pricing worlds under one roof. Clinical Dx is a volume-metered SaaS subscription; the pharma/life-sciences business (OmniScreen, Virchow/PRISM, trial services) is pure custom contract. The same foundation models are simultaneously open-sourced for research and licensed commercially — a deliberate split between a free research wedge and bespoke enterprise monetization.

Strengths & weaknesses

Billing UX : Paige AI billing controls and transparency

• Billing controls — Enterprise contract-based: annual SaaS subscriptions on the clinical side, bespoke licensing/services on the pharma side. There is no in-app plan picker or self-serve upgrade; seat/volume and module changes go through the account team and contract amendments.
• Usage visibility — The clinical product surfaces AI outputs inside the FullFocus viewer and FullFolio image manager, giving labs operational visibility into cases processed, but there is no public cost calculator or published rate card — a lab cannot model spend without a sales conversation and an LIS-integration scope.
• Payment options — Invoiced enterprise contracts via direct sales and a network of digital-pathology partners (Sectra, Roche, Indica Labs, PathPresenter, OptraSCAN, and others). The one-time LIS-integration fee and optional cloud storage are typically separate line items.

Strategic wins : Why Paige AI’s pricing decisions worked

1. Pricing on a metric labs already trust

Anchoring on the slide lets Paige sell into pathology budgets that are already organized by slides and cases, sidestepping the “what’s a token?” education problem pure-usage AI tools face. The FDA authorization lets that slide be billed as a clinical read rather than a research credit. See choosing the right usage metric.

2. Regulatory moat as pricing power

Being the first FDA-authorized AI in pathology (plus 13 CE-IVDD and 13 UKCA marks) turns compliance into a premium: hospitals can use Paige in primary diagnosis, which justifies a sales-led, integration-gated price that unregulated competitors can’t command. Related: how AI companies structure pricing.

3. Two revenue engines from one data asset

The same ~7M-slide dataset and Virchow/PRISM models power both per-slide clinical Dx and custom pharma deals — open-sourcing the models for research seeds a funnel while bespoke licensing monetizes pharma. That dual engine is precisely what made Paige attractive to Tempus. See outcome-based pricing trends.

Areas to improve : Gaps in Paige AI’s pricing approach

1. Zero pricing transparency

A 404 on /pricing and no published methodology force every lab into a sales motion just to learn rough cost — the only public per-slide figure comes from a UK health-technology assessor, not Paige. Even a posted “from £1/slide” or a public integration-fee range would shorten cycles. See bill shock and cost unpredictability.

2. Opaque per-case economics

Because the per-slide rate multiplies with biopsies-per-slide and slides-per-case, the headline ~£1/slide understates real cost, and the promised per-case model is still “in development.” Publishing representative per-case examples would let labs budget honestly instead of discovering the multipliers at quote time.

3. High integration floor blocks small labs

A one-time ~£15,000 LIS-integration fee filters out smaller and community labs that could benefit most from AI-assisted detection. A lighter-weight or partner-subsidized integration tier would widen adoption without undermining the enterprise motion.

Key takeaways

1. Paige publishes nothing — pricing is fully sales-led. The /pricing URL 404s; the only public figures (~£1/slide, ~£15,000 integration) come from the UK’s NICE briefing, not Paige.
2. The value metric is the slide, moving toward the case. Clinical AI is a per-slide SaaS subscription that rises with biopsy volume, slides per case, and storage; a per-case model is in development.
3. Usage meter + enterprise gate. A low-looking per-slide rate travels with a one-time LIS-integration fee and an annual contract — the integration is the real barrier.
4. Two pricing worlds. Clinical Dx is metered SaaS; pharma/life-sciences (OmniScreen, Virchow/PRISM, trial services) is pure custom contract, with the same models open-sourced for research.
5. Regulation is the pricing power — but not enough alone. First FDA authorization justifies a premium, yet Paige sold to Tempus for $81.25M against $220M+ raised, a reminder that a regulatory moat doesn’t guarantee standalone economics.

UBP implications

1. A regulated act can be a usage meter. Paige bills per slide, but each slide is an FDA-cleared diagnostic read — showing that usage-based pricing can attach to clinical events, not just API calls, when regulation underwrites the unit. See usage-based pricing strategy.
2. Couple the meter to the integration. Pairing a per-slide rate with a one-time LIS-integration fee qualifies buyers and funds onboarding — a pattern for any usage-based product that requires deep workflow integration before the meter can run.
3. Split free research from paid enterprise. Open-sourcing Virchow/PRISM for researchers while custom-contracting pharma shows how a foundation-model company can run a free wedge and bespoke monetization off the same asset. See choosing the right usage metric.

Sources

• Paige Diagnostic AI (clinical products) (live capture, accessed 2026-06-10)
• Paige AI Technology & Services (pharma/life sciences) (live capture, accessed 2026-06-10)
• Paige /pricing URL — returns 404 (accessed 2026-06-10)
• NICE Medtech innovation briefing MIB280 — Paige Prostate, “The technology” (per-slide and integration pricing) (second-source pricing, accessed 2026-06-10)
• FDA authorizes software that can help identify prostate cancer (Paige Prostate, De Novo, Sept 2021) (accessed 2026-06-10)
• Tempus AI acquires Paige for $81.25M (Aug 2025) (accessed 2026-06-10)
• Tempus announces the acquisition of Paige (press release) (accessed 2026-06-10)
• Microsoft + Paige train cancer-pathology foundation models (Virchow) (accessed 2026-06-10)

Bottom line

Paige is the first company to win FDA authorization for an AI product in pathology, and it monetizes that lead with a fully sales-led model: clinical diagnostic AI sold as a per-slide SaaS subscription (NICE reports ~£1/slide, rising with volume) on top of a one-time ~£15,000 LIS-integration fee and optional cloud storage, while pharma/life-sciences foundation-model and biomarker work (OmniScreen, Virchow/PRISM) is custom-contracted. There is no public price list — the /pricing URL 404s, and the only per-slide reference comes from the UK’s NICE briefing. Despite raising $220M+ and co-building Virchow with Microsoft, Paige sold to Tempus AI for just $81.25M in August 2025, a reminder that a regulatory moat doesn’t guarantee standalone economics. Browse the pricing blueprint for more fully-researched company profiles.

Want to compare Paige AI against other healthcare and AI-platform companies? Browse the pricing blueprint.