OpenEvidence — AI Research Brief

Executive Summary

Company Overview

Executive Summary

Dominant market position: Over 40% of U.S. physicians use OpenEvidence daily — reportedly more than 500,000 verified clinicians — making it the most widely adopted clinical AI tool in the country by a significant margin.
Rapidly expanding product suite: OpenEvidence has moved well beyond evidence search. Its current platform includes core Q&A search, DeepConsult (asynchronous agentic deep research), Visits (ambient note documentation), and Doctor Dialer (telemedicine). It is building toward a full clinical operating system.
EHR integration is nascent but accelerating: Early enterprise deployments at Mount Sinai and Sutter Health (both via Epic FHIR integration) represent the beginning of a critical transition from standalone app to embedded workflow tool.
HIPAA compliant and SOC 2 Type II certified as of 2025; no FDA clearance required for its current advisory/informational CDS posture, but regulatory classification risk grows as clinical functionality deepens.
Business model creates real conflict-of-interest concerns: OpenEvidence's primary revenue stream is pharmaceutical and medical device advertising at premium CPMs. While ads are siloed from AI-generated answers, this model raises legitimate questions about editorial independence and perception in institutional procurement conversations.
Explosive funding trajectory: ~$735M raised in under 12 months, valuation jumped from $1B (Feb 2025) to $12B (Jan 2026). The company is clearly in a land-grab phase and is investing heavily in the enterprise channel.

Recommendation EXPLORE — Highest physician adoption in the country, maturing EHR integrations, and a free-to-use model make this worth a structured evaluation and demo conversation, with clear eyes on the ad-model tension.

Founding & Background

OpenEvidence was founded in 2021–2022 by Daniel Nadler, PhD, an entrepreneur and Harvard-trained academic with prior experience building AI-driven financial analytics platforms (Kensho, acquired by S&P Global for ~$550M). Nadler founded OpenEvidence after observing that practicing physicians faced an overwhelming and accelerating volume of peer-reviewed literature — reportedly over 1.2 million new studies per year — with no scalable tool to synthesize it at the point of care.

The company is headquartered in New York City. Nadler was named to TIME's TIME100 Health 2025 list. The company is frequently described in the press as "ChatGPT for doctors," a framing Nadler has embraced while emphasizing the platform's strict grounding in peer-reviewed evidence rather than general web knowledge.

Funding Timeline

Feb 2025

$75M round led by Sequoia Capital at $1B valuation — first institutional signal of scale.

Jul 2025

$210M Series B at $3.5B valuation — simultaneous launch of DeepConsult AI agent product.

Oct 2025

$200M round at $6B valuation — continued growth; transition toward enterprise B2B channel.

Jan 2026

$250M Series D at $12B valuation — company announced goal of building "medical superintelligence for doctors." Total raised: ~$735M+.

Key Metrics

40%+ of U.S. physicians use daily | 18M+ clinical consultations/month (Dec 2025) | $100M ARR (as of Jan 2026) | 10,000+ hospitals & medical centers | $12B valuation (Jan 2026) | 100M+ Americans treated by OpenEvidence-using physician (2025)

AI Technology

AI Capabilities & Technology

Model Architecture

OpenEvidence does not publicly disclose its underlying model stack in detail. However, the company is understood to use a combination of proprietary fine-tuned large language models and retrieval-augmented generation (RAG) techniques — pulling answers from a curated corpus of licensed, peer-reviewed medical literature rather than the open web. The company has described its approach as building models trained only on peer-reviewed journals and clinical guidelines.

Notably, OpenEvidence claims its model was not connected to the public internet during training, which the company says substantially reduces hallucination risk by eliminating noisy or low-quality web content from the training signal. OpenEvidence has licensing agreements with leading medical publishers, including NEJM, JAMA, The Lancet, Cochrane, NCCN, and 300+ other peer-reviewed journals, as well as regulatory sources (FDA, CDC).

USMLE Benchmark

OpenEvidence announced that its AI was the first in history to score a perfect 100% on the United States Medical Licensing Examination (USMLE) — a significant benchmark claim that differentiates it from general-purpose LLMs and from prior clinical AI systems. This claim has not yet been independently peer-reviewed as a formal study.

DeepConsult Agentic Architecture

DeepConsult represents OpenEvidence's move into agentic AI. The product uses multiple specialized AI agents that work in parallel to autonomously analyze hundreds of peer-reviewed studies, cross-reference findings, and produce synthesized research reports. The system is designed to operate asynchronously — a physician submits a complex clinical question, and the agents complete the research over a period of minutes to hours, surfacing cross-cutting connections across the literature that would require extensive manual review.

Clinical Validation

A peer-reviewed study published in PLOS ONE (2025) evaluated OpenEvidence's performance against primary care physician clinical decision-making across five complex patient cases involving common chronic conditions. The study found that OpenEvidence consistently provided accurate, evidence-based responses that aligned with physician clinical decision-making, reinforced clinical plans, and provided strong citation support. The study did not find cases where the AI modified (positively or negatively) clinical decisions in this sample. Independent validation at scale is still limited.

Products

Product Suite & Clinical Workflows

OpenEvidence has expanded from a single evidence search tool into a multi-product clinical platform. All products are currently free to verified U.S. healthcare professionals.

Core Evidence Search

Natural-language clinical Q&A grounded in 300+ peer-reviewed journals, FDA, CDC, and major guidelines (NCCN, Cochrane, ACC/AHA). Responses include citations and evidence strength ratings. The foundational product — the entry point for most physicians.

DeepConsult

Asynchronous multi-agent research: a physician submits a complex question and a team of PhD-level AI agents analyzes hundreds of studies in parallel, producing a synthesized, cross-referenced brief. Designed for complex or rare cases where immediate search is insufficient.

Visits

Ambient note documentation that transcribes encounters and enriches the assessment and plan with real-time guideline recommendations and citations. Supports custom physician templates. Positions OpenEvidence to compete directly with ambient scribes like DAX Copilot and Nabla.

Doctor Dialer

Privacy-centric telemedicine platform integrating voice/video calls, messaging, and voicemail with live clinical decision AI embedded in the interface. Positions OpenEvidence as a full clinical communication layer, not just a reference tool.

Clinical Calculators

Evidence-linked calculators (CHADS-VASc, Wells Score, etc.) integrated directly within the search experience, with citations connecting score interpretation back to underlying studies.

CME Credits

Integrated continuing medical education credit tracking — a clinically credible user retention mechanism that incentivizes regular use and deepens platform stickiness for individual physicians.

Target Users and Settings

OpenEvidence is primarily clinician-facing, targeting physicians across all specialties. Mount Sinai's enterprise deployment has expanded access to registered nurses and pharmacists, indicating a broadening to the full clinical care team. The platform serves both outpatient and inpatient contexts, though its search-first model makes it more naturally suited to workflows with even brief moments of physician discretionary time.

Integration

Integration & Technical Architecture

EHR Integration Status

OpenEvidence has historically operated as a standalone web and mobile app, accessed independently of EHR workflows. This allowed it to scale rapidly without the friction of EHR procurement, but capped its workflow integration depth. As of early 2026, the company has begun pursuing formal EHR integrations:

Epic (FHIR-based): Mount Sinai Health System (April 2026) and Sutter Health (February 2026) both announced enterprise deployments embedding OpenEvidence directly within Epic, using FHIR-based integration to allow natural-language evidence searches without leaving the EHR charting environment. Epic App Orchard listing status is not yet publicly confirmed.
SMART on FHIR: The integration pathway leverages SMART on FHIR standards, consistent with Epic's preferred integration model. This allows context-aware queries — theoretically, the EHR patient context could inform the search, though public details on the depth of context passing are limited.
Cerner / Oracle Health: No announced integration as of this writing. This is a meaningful gap for health systems on Cerner.
Meditech, eClinicalWorks, Athena: Not announced.

Mobile Access

OpenEvidence is available as a native iOS app (App Store) and Android app (Google Play), reflecting its origins as a physician point-of-care mobile tool. The mobile experience remains central to its use case for physicians at the bedside or between patient visits.

Data Architecture

OpenEvidence uses edge encryption for PHI handling and states it does not retain patient health information on its servers. Per its privacy policy, the company does collect physician usage data — including specialty, query topics, and engagement patterns — which is used for advertising targeting. Data flows to third-party advertising partners should be clarified in any enterprise BAA negotiation.

Implementation Complexity

For individual physician use: essentially zero — sign up, verify credentials, begin using immediately via web or mobile. For enterprise Epic integration: complexity is higher and mirrors any FHIR-based Epic app deployment. Mount Sinai's rollout represents a reference enterprise deployment but detailed implementation timelines have not been publicly disclosed.

Compliance

Compliance & Security

HIPAA Compliance

OpenEvidence achieved full HIPAA compliance in April 2025, after operating in a pre-PHI mode that restricted physicians from uploading identifiable patient data. The announcement explicitly enabled clinicians to upload PHI securely within the platform. This was a meaningful maturation milestone that removed a prior barrier to enterprise adoption and institutional use.

FDA Regulatory Posture

OpenEvidence's current CDS functionality operates under the FDA's Software as a Medical Device (SaMD) framework in a manner that likely keeps it within the exemption for "decision support software" that displays or communicates clinically-validated information, where a qualified healthcare professional is in the loop interpreting the output. However, as the platform's Visits and DeepConsult features become more prescriptive — e.g., generating treatment plans or making differential diagnoses — the FDA regulatory classification exposure increases. The company has not disclosed any formal FDA engagement, clearance, or De Novo authorization. This is worth probing directly with the vendor.

Data Retention and Privacy

Per its published privacy policy, OpenEvidence collects physician interaction data (search queries, specialty, engagement) which may be shared with advertising partners for targeting. Institutional buyers should scrutinize the BAA to ensure PHI-related data is cleanly segregated from the advertising data layer. Data retention and deletion policies should be explicitly addressed in any enterprise contract.

Pricing

Pricing & Business Model

Individual / Direct-to-Physician

OpenEvidence is entirely free to all verified U.S. healthcare professionals, including all product features: core search, DeepConsult, Visits (ambient documentation), clinical calculators, and CME credits. Physician identity verification is required. The company monetizes individual access through pharmaceutical and medical device advertising, commanding CPMs of $70–$150+ (and reportedly up to $1,000+ for high-value specialist audiences). This is significantly above typical digital ad rates and reflects the platform's unique access to verified prescribers at the point of clinical decision-making.

Enterprise / Health System Channel

OpenEvidence has launched an enterprise channel targeting health systems. Pricing model details are not publicly disclosed, but analyst reporting indicates a per-seat SaaS subscription model for enterprise deployments. Enterprise pricing is estimated to represent a 5–10x ARPU uplift compared to advertising-supported individual access, making it a key growth lever. Enterprise contracts likely include BAA coverage, IT integration support, and administrative provisioning for large clinical teams.

Revenue Metrics

OpenEvidence reached $100M ARR as of January 2026, driven primarily by advertising revenue. The enterprise channel is positioned as the next major revenue driver. The company scaled to this ARR faster than arguably any prior healthcare AI company.

Total Cost of Ownership Considerations

For health systems piloting the free tier: cost is effectively zero beyond IT security review and BAA execution. For enterprise deployments with Epic integration: standard Epic App Orchard integration costs apply, plus per-seat subscription fees. The implicit cost is the advertising model — physician attention is monetized on behalf of pharma and device advertisers, which may create internal perception challenges in academic medical centers focused on evidence independence.

Customer Evidence

Named Health System Customers

Mount Sinai Health System (New York): Announced April 2026 — enterprise-wide deployment embedded within Epic. The first OpenEvidence deployment explicitly extending access to nurses and pharmacists in addition to physicians. This is the marquee enterprise reference customer and signals institutional validation of the platform beyond individual physician use.
Sutter Health (California): Announced February 2026 — Epic-integrated deployment allowing clinicians to search OpenEvidence from within the EHR charting environment.

Broad Physician Adoption

Beyond formal enterprise contracts, OpenEvidence claims over 40% of U.S. physicians use the platform daily — across more than 10,000 hospitals and medical centers. This bottom-up adoption dynamic is unusual in healthcare IT: OpenEvidence achieved massive end-user penetration before formal health system procurement, inverting the typical enterprise sales motion. This creates both an opportunity (institutional buyers have physicians who are already users and advocates) and a compliance risk (physicians may have been sharing PHI through personal accounts before HIPAA compliance was achieved).

Claimed Outcomes

Physicians report spending 40% less time documenting diagnoses and treatments when using OpenEvidence (company-reported).
18 million clinical consultations supported in December 2025 alone (company-reported).
Peer-reviewed study (PLOS ONE, 2025) found performance comparable to physician decision-making in 5 primary care cases.
Independent KLAS rating: Not yet listed as of this research date. This is a notable gap for health system procurement due diligence.

Competitive Landscape

Vendor	Primary Strength	EHR Integration	Pricing Model	Relative Positioning
OpenEvidence	Evidence synthesis + agentic research; dominant U.S. physician adoption	Epic (FHIR, early stage)	Free (ad-supported); enterprise per-seat	Market leader in physician CDS by adoption
UpToDate (Wolters Kluwer)	Decades of trusted, expert-curated content; deep clinical credibility	Epic, Cerner, many EHRs	Institutional subscription ($$$)	Incumbent; strong in institutional settings, weaker in AI-native features
Glass Health	Differential diagnosis + ambient scribe + EHR documentation push	Epic, eClinicalWorks, Athena (SMART on FHIR)	Free tier; $20–$200/mo paid tiers	Strongest workflow integration; less evidence depth
AMBOSS	Medical knowledge library; strong in resident/trainee market	Limited EHR integration	Subscription (individual and institutional)	Trusted for education and exam prep; less point-of-care AI
Epic (native AI)	Embedded natively in EHR; patient context; zero friction	Native Epic	Bundled with Epic platform	Long-term existential threat if Epic expands AI CDS natively
Doximity	Physician social/professional network; clinical messaging	Limited	Free (pharma-sponsored); enterprise	Competitor for physician attention; currently in litigation with OpenEvidence

OpenEvidence's Stated Differentiation

Strict grounding in licensed peer-reviewed literature (vs. general LLMs susceptible to hallucination)
Unmatched physician adoption as network effect / social proof
Free-to-use model removing individual cost barriers
USMLE perfect score benchmark as clinical accuracy signal
Expanding agentic capabilities (DeepConsult) with no comparable product among direct competitors

Red Flags

Red Flags & Open Questions

Pharma Advertising as Primary Revenue Source

OpenEvidence's free tier is monetized through pharmaceutical and medical device advertising at premium CPMs. While the company states that ads are displayed separately from AI-generated answers, this model creates a real (or perceived) conflict of interest — particularly for academic medical centers or systems with strict policies around commercial influence on clinical decision-making. Institutional buyers should ask: What controls prevent advertiser influence on model training or content curation? What are the ad targeting parameters and can they be disabled for enterprise users?

Active Litigation with Doximity

OpenEvidence filed suit against Doximity in 2025, alleging that Doximity impersonated physicians to steal trade secrets. Doximity has countered with its own allegations regarding mishandling of sensitive information and contested HIPAA compliance claims. Active litigation with a major healthcare IT incumbent introduces execution risk, potential distraction for leadership, and reputational uncertainty. Health systems should monitor this case's trajectory.

Pre-HIPAA PHI Exposure Risk

OpenEvidence was not formally HIPAA compliant until April 2025. Given its claimed 40%+ physician adoption rate, it is highly likely that physicians at your institution were using OpenEvidence with some degree of patient context discussion prior to HIPAA compliance. Institutions should assess whether this creates historical exposure and ensure that the current BAA cleanly covers all institutional usage going forward.

FDA Regulatory Classification Risk as Features Deepen

OpenEvidence's current posture likely keeps it within the "decision support" software exemption from FDA clearance requirements. However, as Visits generates evidence-linked treatment plans and DeepConsult produces synthesized clinical recommendations, the boundary with regulated SaMD becomes less clear. No FDA engagement has been publicly disclosed. As you deepen reliance on this platform institutionally, regulatory classification should be probed directly.

Limited Independent Clinical Validation at Scale

The most rigorous published study evaluated only 5 patient cases in primary care — a small sample that cannot support broad claims about safety and accuracy across specialties, rare diseases, or high-acuity settings. Independent KLAS rating does not yet exist. Vendor-reported outcome metrics (40% documentation time savings) lack published methodology. Institutions should request any available formal validation studies and should plan their own internal accuracy assessments before broad deployment.

Epic as Long-Term Platform Risk

Epic's own AI roadmap (including native ambient scribing and embedded CDS) represents a structural long-term competitive threat. There are reported tensions between Epic and OpenEvidence around workflow positioning. If Epic decides to natively bundle CDS capabilities comparable to OpenEvidence, the standalone value proposition diminishes. The FHIR-based Epic integrations are a positive development, but health systems should consider whether this is a partnership or a temporary co-existence before Epic builds it natively.

Key Questions to Ask OpenEvidence in a Demo / Procurement Conversation

What specific advertiser controls exist at the enterprise contract level — can pharma/device ads be disabled for our clinical users?
How is physician query data used, and what data flows to advertising partners? Is this fully blocked under our BAA?
What is your FDA regulatory strategy as Visits and DeepConsult generate more prescriptive clinical outputs?
Can you provide the full Epic App Orchard listing and SMART on FHIR technical documentation?
What is your implementation timeline and FTE burden for an Epic-embedded enterprise deployment?
Do you have any published accuracy benchmarks broken down by specialty or acuity level?
What is the status of the Doximity litigation and what is your legal team's projected timeline for resolution?

Resources