Yardstick Research tear-sheet / AI sales cohort

Methodology · how we score · rubric weights in plain sight · vendors received this sheet seven days before publication and could flag factual errors, never rankings

Openevidence

Identity

• Founded: September 2021 by Daniel Nadler, PhD (Founder + CEO; previously founded Kensho, sold to S&P Global for $550M) and Zachary Ziegler (co-founder, ML). [VENDOR-CLAIMED + THIRD-PARTY corroboration - https://research.contrary.com/company/openevidence, https://en.wikipedia.org/wiki/OpenEvidence]
• HQ: Miami, FL per Wikipedia + Crunchbase News + recent press cycles; Cambridge, MA per Sacra equity-research profile. Treat Miami as the operating HQ and surface the Cambridge entry as worth confirming during right-of-reply. [THIRD-PARTY - https://en.wikipedia.org/wiki/OpenEvidence, https://sacra.com/c/openevidence/]
• Legal entity: OpenEvidence Inc.
• Funding: ~$700M total raised across Series A through Series D. Most recent: $250M Series D at $12B post-money valuation, co-led by Thrive Capital and DST Global, announced January 21, 2026. Series C: $200M at $6B led by GV (October 2025). Series B: $210M at $3.5B co-led by GV and Kleiner Perkins (July 2025). Series A: $75M at ~$1B led by Sequoia (February 2025). [THIRD-PARTY - https://techcrunch.com/2026/01/21/openevidence-hits-12b-valuation-with-new-round-led-by-thrive-dst/, https://news.crunchbase.com/venture/openevidence-ai-doctors-doubles-valuation-seriesd/, https://en.wikipedia.org/wiki/OpenEvidence]
• Headcount: [UNKNOWN] No third-party headcount surfaced across Sacra, Contrary Research, Crunchbase News, TechCrunch, or Wikipedia at research date.
• Cap-table notables: Thrive Capital, DST Global, GV, Sequoia, Kleiner Perkins, Coatue, Blackstone, Bond, Craft Ventures, Nvidia, Mayo Clinic. Mayo Clinic is both a Platform Accelerate program partner and a cap-table investor - an unusual buyer / investor overlap worth surfacing during procurement.
• Leadership:
• Daniel Nadler, PhD - Founder + CEO. TIME100 Health 2025.
• Zachary Ziegler - Co-founder + CTO. Harvard ML background.
• [VENDOR-CLAIMED + THIRD-PARTY - https://research.contrary.com/company/openevidence]
• Recent news (last 12 months):
• April 2026 - Mount Sinai Health System enterprise-wide Epic integration across seven hospitals, extending access to physicians + nurses + pharmacists. [THIRD-PARTY - https://www.mountsinai.org/about/newsroom/2026/mount-sinai-health-system-collaborates-with-openevidence-to-provide-evidence-based-knowledge-within-electronic-medical-record]
• March 10, 2026 - 1 million clinical consultations between NPI-verified physicians and the AI system in a single 24-hour period. [VENDOR-CLAIMED + THIRD-PARTY - https://www.prnewswire.com/news-releases/openevidence-achieves-historic-milestone-1-million-clinical-consultations-between-verified-doctors-and-an-artificial-intelligence-system-in-a-single-day-302712459.html]
• March 2026 - Coding Intelligence launched: inline ICD-10, E/M, and CPT code suggestions. [VENDOR-CLAIMED - https://www.fiercehealthcare.com/health-tech/openevidence-rolls-out-ai-medical-coding-feature]
• February 11, 2026 - Sutter Health Epic integration via FHIR-based architecture launched. [THIRD-PARTY - https://www.businesswire.com/news/home/20260211318919/en/Sutter-Health-Collaborates-with-OpenEvidence-to-Bring-Evidence-Based-AI-Powered-Insights-into-Physician-Workflows]
• January 21, 2026 - Series D $250M at $12B valuation co-led by Thrive Capital + DST Global. [THIRD-PARTY - https://techcrunch.com/2026/01/21/openevidence-hits-12b-valuation-with-new-round-led-by-thrive-dst/]
• November 29, 2025 - First independent peer-review-tracked accuracy benchmark posted to medRxiv: subspecialty pilot scored Quick Consult at 34% and DeepConsult at 41% accuracy on 100 MedXpertQA scenarios. Preprint, not yet peer-reviewed. [THIRD-PARTY - https://www.medrxiv.org/content/10.64898/2025.11.29.25341091v1]
• August 15, 2025 - First AI to score 100% on a USMLE administration. [VENDOR-CLAIMED - https://www.prnewswire.com/news-releases/openevidence-creates-the-first-ai-in-history-to-score-a-perfect-100-on-the-united-states-medical-licensing-examination-usmle-302531156.html]
• August 2025 - Visits launched: ambient transcription of patient encounters with assessment-and-plan enrichment. [VENDOR-CLAIMED - https://www.openevidence.com/announcements/visits-real-time-medical-intelligence]
• July 15, 2025 - DeepConsult wide release: autonomous multi-step reasoning agent for physicians. [VENDOR-CLAIMED - https://www.openevidence.com/announcements/openevidence-the-fastest-growing-application-for-physicians-in-history-announces-dollar210-million-round-at-dollar35-billion-valuation]
• June 2025 - JAMA Network multi-year content agreement covering 12 journals. [VENDOR-CLAIMED - https://www.fiercehealthcare.com/ai-and-machine-learning/jama-signs-multi-year-deal-openevidence-inform-ai-powered-medical-search]
• April 2025 - HIPAA compliance achieved; CME credit program launched. [VENDOR-CLAIMED - https://research.contrary.com/company/openevidence]
• February 2025 - NEJM Group multi-year content agreement (all content from 1990 forward). [VENDOR-CLAIMED - https://www.openevidence.com/announcements/openevidence-and-nejm]
• Latent risk: Ad-supported clinical decision support is structurally novel for the cohort - pharmaceutical + medical-device CPMs at $70-$1000+ inside a tool physicians use at the point of care. Vendor states ads "cannot influence answers" but does not publicly disclose the architectural wall between ad-targeting metadata and clinical-question / Visits / PHI data. CEO Daniel Nadler has signaled the model "may change" but no pivot announced. Independent commentary (the-geyser.com, Sacra) flags this as a procurement-due-diligence question with sharply varying buyer reactions. [THIRD-PARTY - https://www.the-geyser.com/openevidence-targets-ads/, https://sacra.com/c/openevidence/]
• Archetype: Physician-facing AI search and reasoning platform anchored on licensed peer-reviewed content (NEJM + JAMA + NCCN + AMA + AAFP + ACEP + Cochrane). Free to every verified US NPI. Materially distinct from the cohort's ambient-scribe vendors (Abridge / Ambience / Nabla / Suki / DAX Copilot) - primary surface is a Google-style citation-grounded search box for the doctor, not a recorded patient conversation. The Visits feature added an ambient-scribe-shaped surface in August 2025 but is not the product center of gravity. Cohort-leading adoption (1M consultations in a single day, 18M monthly), cohort-leading citation provenance, cohort-trailing standardized-clinical-benchmark transparency on subspecialty reasoning.

Total score: 76.25 / 100

• Stage fit:
• Foundation (<40 readiness): yes - The "free, no procurement, no Epic project required" surface is uniquely valuable at this stage. Pull-through adoption is bottom-up via the physician's own NPI; a Foundation-stage health system can validate the product with zero IT cost.
• Pilot (40-59 readiness): yes - Citation-grounded evidence search + DeepConsult research depth fit the Pilot-stage pattern of "build a use case, measure satisfaction, justify the next stage." Visits is the natural Pilot-stage surface to pair against an ambient-scribe pilot.
• Scale (60-79 readiness): conditional - Strong on adoption + content + cost; conditional on the buyer's tolerance for the EHR-integration-depth gap vs Abridge + Ambience. Sutter + Mount Sinai are the canonical Scale-stage Epic-embed pattern.
• Optimization (80+ readiness): conditional - Strong on clinical-evidence-search depth; conditional on Epic integration depth + multi-EHR breadth. Optimization-stage buyers typically run multiple major EHRs; OpenEvidence's Cerner / Meditech / Athena depth is undocumented at research date.
• One-line verdict: The cohort's adoption + citation-provenance leader and the only major vendor offering full functionality free to every verified US physician - constrained by a USMLE-vs-subspecialty accuracy gap (100% vs 34-41%), shallow non-Epic EHR depth, missing HITRUST disclosure, and a structurally novel ad-supported clinical-decision-support model.

Headline numbers

Dimension scores

Pricing detail

Source: Reconstructed from vendor announcements + Sacra equity-research profile + Contrary Research. OpenEvidence does not publish a public enterprise pricing page. - https://sacra.com/c/openevidence/, https://research.contrary.com/company/openevidence

• Consumer tier - $0 for every verified NPI-holding US clinician. [VENDOR-CLAIMED] Full DeepConsult + Visits + Coding Intelligence + AI Dialer + CME + unlimited search across the licensed content corpus. Solo physicians, residents, mid-size groups, and FQHC clinicians all access at zero cost. - https://www.openevidence.com/announcements/visits-real-time-medical-intelligence
• Implicit ad-supported consumer-tier ARPU ~$124 per verified user per year. [THIRD-PARTY] Sacra calculation: $150M FY 2025 revenue / ~1.2M verified-user-equivalent base. Pharmaceutical + medical-device advertising at $70-$1000+ CPMs (vs $5-$15 typical of consumer social platforms) is the monetization. - https://sacra.com/c/openevidence/
• Enterprise EHR-embedded tier - undisclosed list, triangulated at 5-10x consumer ARPU. [THIRD-PARTY] Sacra describes Sutter + Mount Sinai as the enterprise-tier proof points with an implied per-clinician-per-year band of ~$620-$1,240 if the consumer ARPU anchor holds. No public rate card. - https://sacra.com/c/openevidence/
• Implementation / integration fees: [UNKNOWN] Not disclosed for either Sutter (Feb 2026) or Mount Sinai (Apr 2026). Worth raising in right-of-reply.
• No paid consumer tier exists. The full product surface is free with NPI verification; no "Pro" or "Premium" upsell on the consumer surface. The only revenue uplift path is enterprise EHR embed.
• Triangulation note: vendor public statement of $100M ARR (January 2026 per Crunchbase News) aligns with the $150M 2025 full-year revenue Sacra reports, implying meaningful ARR run-rate growth Q4 2025 into Q1 2026. Both numbers are consistent with the December 2025 ~6x year-over-year monthly-consultation growth (~3M to 18M).

Integrations

• Epic (the primary integration anchor): [THIRD-PARTY] Two named enterprise deployments at research date.
• Sutter Health (California) - announced February 11, 2026. FHIR-based architecture; evidence search inside Epic workflows at the point of care. - https://www.businesswire.com/news/home/20260211318919/en/Sutter-Health-Collaborates-with-OpenEvidence-to-Bring-Evidence-Based-AI-Powered-Insights-into-Physician-Workflows
• Mount Sinai Health System (NY) - announced April 2026. Enterprise-wide deployment across seven hospitals; OpenEvidence's first deal to extend access across the full clinical care team (physicians + registered nurses + pharmacists). - https://www.mountsinai.org/about/newsroom/2026/mount-sinai-health-system-collaborates-with-openevidence-to-provide-evidence-based-knowledge-within-electronic-medical-record, https://hitconsultant.net/2026/04/01/mount-sinai-openevidence-ai-epic-integration-nurses-pharmacists/
• Epic Partners and Pals tier: [UNKNOWN] Not confirmed at any tier. Abridge holds first-Pal status; OpenEvidence is not on the publicly enumerated Partners-and-Pals list at research date.
• Epic Showroom listing: [UNKNOWN] Not surfaced.
• Cerner / Oracle Health CODE program: [UNKNOWN] No deployment confirmed. Contrary Research mentions SMART on FHIR could enable Cerner integration but no current Cerner deployment is on the public surface. - https://research.contrary.com/company/openevidence
• Meditech / Athenahealth / eClinicalWorks marketplace listings: [UNKNOWN] None surfaced as first-class integrations.
• SMART on FHIR + FHIR R4 conformance statement: [UNKNOWN] No published conformance statement.
• Developer surface (API / SDKs / sandbox): [UNKNOWN] No public developer portal; integration model is "free standalone consumer app + selective Epic embed on enterprise contract."
• Standalone consumer-tier app: [VENDOR-CLAIMED] Web + mobile, free for verified US NPIs. The dominant adoption pattern. - https://www.openevidence.com/
• Content partnerships (licensed corpus, not connector integrations): NEJM Group (all content from 1990 forward), JAMA Network (12 journals), NCCN, AMA, AAFP, ACEP, Cochrane Library, PubMed, FDA, CDC. - https://www.openevidence.com/announcements/openevidence-and-nejm, https://www.fiercehealthcare.com/ai-and-machine-learning/jama-signs-multi-year-deal-openevidence-inform-ai-powered-medical-search
• RCM / clearinghouse connectors: [UNKNOWN] Coding Intelligence (Mar 2026) surfaces ICD-10 / E/M / CPT suggestions inline in the documentation workflow; specific Waystar / Change Healthcare / Optum / Availity / Epic Resolute connectors are not enumerated on any public integrations page.
• HIE / TEFCA QHIN participation: [UNKNOWN] Not surfaced.

Editorial assessment

OpenEvidence is the most-adopted physician-facing AI product in US healthcare by every consultation-volume metric available, and it occupies a structurally distinct seat from the rest of the healthcare-clinical cohort. Where Abridge, Ambience, Nabla, Suki, and DAX Copilot compete on Epic-native ambient-scribe documentation depth, OpenEvidence's product center of gravity is a citation-grounded clinical-evidence search engine that any verified US physician can use for free, anywhere, with no IT integration required. The March 10, 2026 milestone of 1 million clinical consultations in a single 24-hour period is the largest physician-side scale signal in the cohort by any metric, and the inline citation surface to NEJM / JAMA / NCCN / Cochrane / AMA-licensed content is the strongest citation provenance in the consumer-physician AI category per the independent iatroX 2026 landscape assessment.

The load-bearing tension in the dossier is the accuracy gap between standardized-exam performance and complex subspecialty reasoning. The vendor's August 2025 USMLE 100% score is real and a meaningful signal of structured-question competence. The November 2025 medRxiv pilot on 100 MedXpertQA subspecialty scenarios scored OpenEvidence Quick Consult at 34% and DeepConsult at 41%, placing OpenEvidence in the middle of the LLM pack (14-46% across 11 LLMs evaluated on the same dataset). Both numbers belong in the buyer's understanding. The preprint is not yet peer-reviewed and the authors disclaim against using it to guide clinical practice, but it is the strongest independent signal on real-world clinical-task performance available at research date. The source-recency case documented in independent commentary - OpenEvidence recommending graded exercise therapy for ME/CFS based on guidance NIH reversed in 2022 - is the operational manifestation: the licensed-content moat is simultaneously a freeze-the-evidence-at-license-date risk that buyers should expect to manage through procurement-side guideline-refresh cadence requirements.

The compliance posture documents the certifications buyers expect (HIPAA, BAA, SOC 2 Type II, encryption at rest and in transit) and leaves three load-bearing questions open: HITRUST CSF status, the model-training opt-out default, and the architectural wall between ad-targeting metadata and clinical-question / Visits / PHI data. The ad-supported clinical-decision-support model is structurally novel in the cohort - pharmaceutical + medical-device CPMs at $70-$1000+ inside a tool physicians use at the point of care. CEO Daniel Nadler has publicly signaled the ad model "may change." Buyer reactions to the model vary sharply: some health systems are comfortable because pharma marketing to physicians is already a $20B / year industry that exists with or without OpenEvidence; others view it as a procurement-disqualifying conflict in a clinical-decision tool. The cohort's other major vendors do not have an analogous structural conflict for procurement to evaluate.

The EHR integration depth is in early-stage build-out relative to the cohort's leaders. Sutter (February 2026) and Mount Sinai (April 2026) are the only two named Epic-embedded enterprise deployments at research date - both meaningful but both early - against a cohort backdrop where Abridge holds first-Pal status in Epic's Partners and Pals program with a 150+ health-system production footprint and Ambience holds Toolbox + Hyperdrive + Haiku depth across multiple major IDNs. No Cerner / Meditech / Athena / eClinicalWorks integration is confirmed at research date. The bet a buyer makes at the enterprise tier is that OpenEvidence's content + adoption + cost-economics advantage is large enough to outweigh the integration-depth gap; that bet is defensible for an Epic-anchored health system and gets harder for a buyer running mixed-EHR or non-Epic primary.

The cost economics are the dossier's brightest spot. The full consumer product surface is free to every verified US physician with no per-seat purchase decision, no procurement cycle, no Epic-integration project, no IT lift. For Foundation- and Pilot-stage buyers the cost ceiling is zero, which makes adoption an evidence-and-policy question rather than a budget question. When to revisit: when OpenEvidence publishes HITRUST CSF tier and audit date, the underlying LLM stack disclosure, a documented model-training opt-out default, an ad-targeting-vs-clinical-data architectural wall statement, the next independent peer-reviewed accuracy benchmark on real-world clinical tasks, or a Cerner / Oracle Health CODE program tier. Any one closes a current gap; two or three would push the case for OpenEvidence as the cohort #1 rather than the strong-second placement the rubric currently produces.

Best for

• Stage: Foundation, Pilot, conditional Scale, conditional Optimization. Strongest at Foundation + Pilot because of the zero-cost consumer-tier adoption surface; conditional at Scale + Optimization because of the EHR-integration-depth gap vs cohort leaders.
• Company profile: Any US-physician-employing organization at any size from solo to enterprise IDN. Strongest fit: Epic-anchored regional or national health systems that want a citation-grounded evidence layer adopted bottom-up at zero cost, with the option to deepen into Epic via the Sutter / Mount Sinai integration pattern. Also strong for FQHC + community health where the per-clinician pricing of cohort scribe vendors is a procurement constraint.
• Industry fit: Strong for hospital / health-system, ambulatory groups, FQHC / community health, nursing + pharmacy (Mount Sinai April 2026 extension), emergency medicine (ACEP is a content partner), solo / small practice (the cohort's only product that works at zero cost for an individual physician). Moderate for ASC / surgical (evidence search applies; ambient-scribe specialists deeper on op-note workflows). Weak for imaging / radiology (out of product surface). Conditional for behavioral health (evidence search applies; ad-architecture + source-recency questions surface harder).
• Sales motion: Bottom-up adoption via individual physician NPI verification. Enterprise sales overlay for Epic-embedded deployments (Sutter, Mount Sinai pattern). No per-seat purchase decision at the consumer tier.
• Annual tooling budget: $0 at the consumer tier for individual physicians. Undisclosed at the enterprise EHR-embedded tier; Sacra triangulates 5-10x consumer ARPU implying roughly $620-$1,240 / clinician / year if the $124 consumer ARPU anchor holds - still well below the cohort's ambient-scribe vendor list rates.
• Skip if: You need (a) a deep Epic-Hyperdrive ambient-scribe with bi-directional write across multiple clinical + orders resources as the procurement bar (Abridge / Ambience are deeper at research date), (b) Cerner / Meditech / Athena / eClinicalWorks as the primary EHR with Epic-deep integration depth as the bar (OpenEvidence is Epic-only on the enterprise surface at research date), (c) published RCT or pragmatic-trial clinician-burden-reduction evidence specific to OpenEvidence (none published at research date), (d) FDA-cleared SaMD depth (OpenEvidence is informational decision support; no SaMD clearance), (e) HITRUST r2 as a procurement gate (not surfaced on public materials), (f) sub-$200 / clinician / month ambient-scribe specialist coverage as the primary procurement criterion (OpenEvidence is an evidence engine first; the Visits scribe surface is shallower than the cohort scribe specialists at research date), or (g) an ad-free clinical-decision-support model as a procurement requirement (OpenEvidence's monetization is pharmaceutical + medical-device advertising at the consumer tier).

Right-of-reply

OpenEvidence received this tear-sheet seven calendar days before publication of the Yardstick Research 2026 Yardstick Report, including all measured numbers, sample outputs, and editorial assessment. OpenEvidence was given the opportunity to flag factual errors - incorrect pricing, misquoted feature availability, outdated screenshots, factual misstatement in the editorial assessment. OpenEvidence was not given the opportunity to request a score revision, dispute the rubric or its weights, withdraw from inclusion, negotiate ranking placement, or suggest changes to the editorial assessment beyond factual correction. Where OpenEvidence flagged a factual correction, the correction was applied if verified and noted here; where OpenEvidence disputed scoring, the dispute is recorded in the appendix but the score stands. Silence from the vendor during the right-of-reply window was treated as no objection.

Right-of-reply gaps

Specific [UNKNOWN] items surfaced in the dossier and explicitly raised with the vendor in right-of-reply:

1. Underlying LLM stack. Is the production inference fully proprietary, or does it orchestrate over frontier providers (OpenAI / Anthropic / Google)? If frontier-provider, which models on which workloads, and where do those API calls land geographically?
2. Model-training opt-out default. Is physician-side query data used to train the specialized models? Is PHI captured in Visits used to train? Is opt-out configurable at the enterprise-contract level? What is the default in the standard MSA / BAA / DPA?
3. Ad-targeting architectural wall. Is the ad-targeting metadata pipeline architecturally independent of the clinical-question / Visits / PHI data pipeline? Can the vendor publish a one-page wall statement?
4. HITRUST CSF certification. Held or not held? If held, which tier (e1 / i1 / r2)? Audit date?
5. SOC 2 Type II audit date. Disclosed certification but not date.
6. ISO 27001, PCI-DSS, FedRAMP, TX-RAMP. Any of these held?
7. MedHELM / MedQA / MedBench / MedConceptsQA performance. Beyond the 100% USMLE score, are there published standardized-benchmark numbers OpenEvidence is willing to publish or co-author?
8. FDA SaMD positioning. With Coding Intelligence + Visits + DeepConsult expanding the product surface, what is the current regulatory positioning and trajectory?
9. FHIR R4 conformance and resource scope. Which FHIR resources does the Sutter / Mount Sinai integration consume + write back? Is there a published conformance statement?
10. Cerner / Oracle Health CODE program participation. Tier? Roadmap?
11. Meditech / Athenahealth / eClinicalWorks marketplace listings. Live? At what depth?
12. HIE / TEFCA QHIN participation. Live?
13. Enterprise per-clinician list rate. Sutter + Mount Sinai contract economics?
14. Sutter + Mount Sinai clinician counts at production. How many clinicians are live on each deployment at research date?
15. Source-recency cadence. How frequently is the indexed-content corpus refreshed against guideline updates (NIH, USPSTF, specialty-society)? What is the SLA for incorporating a published reversal of prior guidance?
16. Sub-processor list. Public list?
17. Headcount. No third-party number surfaced as of research date.
18. HQ confirmation. Miami, FL (per Wikipedia + Crunchbase News + recent press) vs Cambridge, MA (per Sacra)?
19. Ad-revenue trajectory. CEO Nadler has signaled the ad model "may change" - timeline? What replaces or supplements it?
20. OpenEvidence Visits scribe depth vs cohort ambient-scribe leaders. EHR-write specifics? How does Visits compare functionally to Abridge Inside on Epic Hyperdrive for the chart-write step?
21. NPI verification depth. Anti-fraud controls beyond NPI lookup at signup?
22. Pricing transparency. Any willingness to publish an enterprise rate-card range?