OpenEvidence
B2B2C ad-supported freemium: free for NPI-verified US clinicians, monetized via pharmaceutical and medical-device advertising at premium CPMs ($70-$1,000+ per Sacra, vs $5-15 for social platforms), plus emerging enterprise/EHR deployments with health systems (a non-ad enterprise tier is in development)
All primary points are reported post-money valuations from press coverage of the rounds; the mid-2026 secondary point is an indicative floor from thin, model-derived platform quotes, not a priced transaction. 12x valuation growth Feb 2025 to Jan 2026.
The thesis on this name
State of AI for Healthcare
PRIVATE — a watchlist name with no listed instrument, and the strongest validation of the corrected thesis. The private twin of the Doximity economics: the payer is pharma's commercial budget, the product is free to the verified clinician, and it holds zero FDA authorisations (a verified zero across all 1,524 rows of the AI-Enabled Medical Device List) and no CPT code. It needs neither, because it never asks a health system's IT budget for anything. Reported revenue is ~$300M annualised as of Jul 2026, roughly double the ~$150M at end-2025 — trade press, with no primary source behind it, because the company publishes nothing. The last disclosed mark is $12B post-money on a $250M Series D co-led by Thrive Capital and DST Global (22 Jan 2026), alongside nearly $700M raised over the trailing twelve months, which is a TTM basis and not a lifetime total. Against that mark the revenue multiple is ~40x on Jul-2026 revenue and ~80x on revenue at the January mark date; both are reported-revenue-implied, not analytical findings. A higher-valuation round reported in July 2026 is unconfirmed and was itself reported unlikely to proceed on dilution grounds; it is not carried as a mark anywhere on this board, and the $12B stands as the last disclosed one. The scale comparison is the board's core structural fact, and it needs its as-of every time: at $12B, one private company is 35.4% of the entire 17-name pure-play cohort (16 US-exchange-listed plus OTC-quoted Veradigm, Nasdaq-delisted 2024) (~$33.90B at the 31 Jul 2026 close) and exceeds Tempus ($7.92B) plus Doximity ($3.76B) combined by $320M — a 2.7% margin that a 3% move in either listed name flips. It stays a watchlist name rather than a position because the only available vehicles mark to the last round rather than to revenue, so they express the primary market's opinion of the thesis instead of the thesis.
Earnings, margins, COGS & capex
Private and unaudited. Company disclosed passing $100M revenue in 2025; Sacra estimates $150M 2025 revenue (up from $7.9M in 2024) with ~90% gross margins. Monetization is Doximity-style pharma advertising against an unusually valuable audience (verified prescribers at the clinical point of decision), yielding ARPU around $124 per Sacra (vs Doximity's ~$228 per clinician). Cost base is model inference/compute, journal content licensing (NEJM Group, JAMA Network, Cochrane, NCCN, Wiley), and engineering/sales headcount. Burn is funded by ~$765M raised since founding; profitability not disclosed.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~10¢ is cost of goods and ~90¢ operating expense, leaving ~0¢ of operating profit.
Revenue trend
Margins
Stable-to-improving as ad inventory scales on fixed content/compute base
Presumed negative; heavy investment phase
Unknown
COGS structure
AI inference compute (cloud GPUs; Nvidia is an investor), medical-journal licensing fees (NEJM Group content from 1990 forward, JAMA Network, Cochrane, NCCN, Wiley's 400+ journals and references), NPI verification and platform hosting. Ad-serving cost is minimal, which underpins the estimated ~90% gross margin.
Capex
Not disclosed. Asset-light software model — compute is rented, not owned; no known data-center capex.
Latest earnings
n/a
None issued; CEO Daniel Nadler publicly stated revenue topped $100M in 2025
- Last priced round
- $250M Series D at $12B post (Jan 2026, co-led by Thrive Capital and DST Global)
- Prior rounds
- $75M Series A at $1B (Feb 2025, Sequoia); $210M Series B at $3.5B (Jul 2025, co-led by GV and Kleiner Perkins); $200M Series C at $6B (Oct 2025, GV)
- Total raised
- ~$765M since founding — $735M across four priced rounds Feb 2025-Jan 2026 plus ~$32M of derived pre-institutional founder/angel capital 2021-2023 (backers also include Coatue, Blackstone, Bond, Craft Ventures, Conviction, Greycroft, Nvidia, Mayo Clinic)
- Consultation volume
- ~20M/month (Jan 2026); 1M in a single day (Mar 10, 2026); ~27M clinical encounters (Apr 2026)
- Physician reach
- >40% of US physicians logging in daily (company claim); 10,000+ hospitals and medical centers; 65,000+ new verified clinicians/month
- Reported ARPU
- ~$124 (Sacra estimate) vs Doximity's ~$228
Growth drivers
- Physician penetration — company claims over 40% of US physicians log in daily across 10,000+ hospitals and medical centers, with 65,000+ new verified US clinicians registering monthly
- Query volume compounding — ~3M consultations/month early 2025 to ~20M/month Jan 2026 and ~27M clinical encounters in Apr 2026, expanding monetizable ad inventory
- Enterprise/EHR channel — system-wide Epic-integrated deployments at Sutter Health (Feb 2026), Mount Sinai (Mar 2026), Cedars-Sinai (May 2026) opening a second revenue leg; a non-ad enterprise pricing tier is in development
- Content-moat expansion — multi-year licenses with NEJM Group (Feb 2025), JAMA Network (Jun 2025), Cochrane, NCCN, and Wiley (400+ journals, Mar 2026) deepen answer quality vs generic LLMs
- Pharma ad budgets migrating from Doximity/Medscape-style channels to point-of-decision AI answers with far higher CPMs
- International expansion and adjacent clinician segments (NPs, PAs, nurses, pharmacists — Mount Sinai deployment already extends beyond physicians) remain largely untapped
Bull & bear
OpenEvidence is compounding into the default clinical-knowledge interface for American medicine — a winner-take-most network with a licensed-content moat, Doximity-grade ad economics at the higher-intent point-of-decision moment, and an untouched enterprise second act.
- Distribution is the moat: company claims >40% of US physicians daily-active with near-zero CAC; at that penetration the platform becomes standard-of-care infrastructure, like UpToDate was for 20 years
- Monetization barely started: ~$124 ARPU (Sacra) vs Doximity's ~$228 on a lower-intent surface — closing that gap alone roughly doubles revenue on flat usage, before ad-load or enterprise pricing kicks in
- Content licenses (NEJM archive from 1990, JAMA Network, Cochrane, NCCN, Wiley's 400+ journals) are a legally enforceable data moat that generic LLMs cannot replicate
- Enterprise EHR deployments (Sutter, Mount Sinai, Cedars-Sinai in four months) prove a channel that, once the planned paid enterprise tier ships, could re-rate the company from ad-comp to healthcare-SaaS-comp multiples
- Usage is still accelerating post-scale: ~3M consultations/month early 2025 to ~20M in Jan 2026, a 1M-consultation single day in Mar 2026, and ~27M encounters in Apr 2026
- Estimated ~90% gross margins mean incremental ad dollars flow almost entirely to contribution; the model turns profitable quickly at scale
- Cap-table quality (Sequoia, GV, Kleiner Perkins, Thrive, DST, Coatue, Blackstone, Nvidia, Mayo Clinic) provides capital, distribution and credibility through any downturn
A $12B mark on roughly $100-150M of unaudited 2025 revenue prices in perfection for a two-front war: an entrenched incumbent above and free frontier models below, with a single ad-revenue engine that regulators and physician trust could break.
- Valuation froth: ~80x Sacra's $150M estimate (~120x the disclosed $100M floor) after a 12x valuation increase in eleven months; the closest public comp, Doximity, trades near 6x TTM sales (~$4B market cap on ~$645M revenue, Jul 2026)
- Frontier-model commoditization: ChatGPT, Claude and Gemini answer medical questions increasingly well for free; if 'good enough' generalist answers win, OpenEvidence's engagement and ad inventory erode
- UpToDate counterattack: Wolters Kluwer has decades of hospital contracts, editorial trust and is shipping genAI features; enterprise buyers may default to the incumbent's AI rather than adopt a challenger
- Ad-model fragility: pharma promotion at the point of clinical decision invites FDA/OIG/FTC scrutiny and academic backlash; a forced separation of ads from answers would gut the CPM premium
- Supplier concentration: NEJM and JAMA licenses are the product's spine; renewal-time pricing power sits with the journals, and they could launch or license competing AI products
- Trust is one incident away: a documented patient harm traced to a hallucinated or stale answer would trigger physician churn, health-system bans and plaintiff litigation
- All key financials (revenue, margins, daily-usage share) are company-sourced or analyst-estimated and unaudited; private-market history is littered with growth stories that shrank under S-1 scrutiny
What it is worth
Last priced round anchored, cross-checked against revenue multiples and public physician-network comps (Doximity at ~6x TTM sales / ~$4B market cap as of Jul 2026; Wolters Kluwer Health / UpToDate as the incumbent yardstick)
$2-5B
ad-model regulatory constraint, UpToDate/frontier-model competition, or a trust incident stalls revenue near $200M and the multiple resets toward public physician-ad-network territory (Doximity trades ~6x sales; even a still-premium 10-25x implies $2-5B) — a 60-80% markdown from the Series D
$10-15B
growth remains strong but the multiple compresses toward 25-40x as revenue reaches $300-400M; company grows into the Series D mark rather than beyond it
$20-30B+ by a 2027-2028 IPO
revenue scales toward $500M-$1B on ARPU catch-up to Doximity's ~$228, ad-load expansion and a paid enterprise tier, and the market awards a premium network multiple (20-30x forward) as the default clinical-knowledge layer
Jan 2026 Series D set $12B post against 2025 revenue of $100M+ (company-disclosed) to ~$150M (Sacra estimate) — roughly 80-120x, an AI-hypergrowth multiple that assumes revenue compounds several-fold and the enterprise leg materializes. Secondary indicative quotes (Hiive ~$462, NPM ~$521, Notice ~$575/share, mid-2026) sit at-to-above the round but are thin and model-derived. All revenue figures are unaudited. Not financial advice.
SWOT
Strengths
- Fastest-growing physician application in history (company claim, largely unchallenged in press): >40% of US physicians claimed as daily users, acquired at near-zero CAC via word of mouth
- Licensed content moat — NEJM Group (all content from 1990 forward), JAMA Network, Cochrane (official AI partner), NCCN, and Wiley (400+ journals) with forced citations — generic LLMs cannot legally reproduce this corpus
- Doximity-proven monetization with better placement — pharma ads at the moment of clinical decision command $70-$1,000+ CPMs (Sacra) vs $5-15 for social platforms
- ~90% estimated gross margins and an asset-light model
- Elite cap table (Sequoia, GV, Kleiner Perkins, Thrive, DST Global, Coatue, Blackstone, Nvidia) plus Mayo Clinic as strategic partner-investor
- Repeat founder: Daniel Nadler previously built and sold Kensho to S&P Global for ~$550M (2018)
Weaknesses
- Single revenue engine — overwhelmingly dependent on pharma/device advertising, a cyclical and heavily regulated budget line
- Free product with no user lock-in contract — physicians can switch to a better answer engine instantly
- Unaudited financials — revenue, margin and daily-usage claims are founder/press/analyst-sourced with no third-party verification
- Journal licenses are renewable multi-year deals, not owned IP — key-supplier concentration in NEJM/JAMA
- Young company (founded 2021) scaling org, trust and safety processes at extreme speed
- Ad-supported clinical answers create an inherent perceived conflict of interest that competitors will weaponize
Opportunities
- Enterprise/EHR-embedded deployments (Sutter, Mount Sinai, Cedars-Sinai already live, currently on the ad-supported model) — a planned paid enterprise tier would add a second revenue leg less exposed to ad cycles
- International clinician markets and non-physician clinicians (NPs, PAs, nurses, pharmacists)
- Adjacent workflows — differential diagnosis support, hands-free voice AI at the bedside, prior-auth documentation, patient-facing evidence summaries
- Pharma budgets beyond display ads: sponsored medical education, launch analytics, real-world evidence insights
- Becoming the default clinical-knowledge layer inside EHRs, displacing UpToDate's roughly $595M franchise (Sacra estimate of UpToDate revenue)
- IPO window: at reported scale and growth, a 2027-2028 listing is plausible, giving late investors an exit path
Threats
- Wolters Kluwer's UpToDate (the entrenched incumbent, embedded in hospital contracts) shipping competitive genAI search
- Frontier-lab generalists — OpenAI, Anthropic, Google — improving medical answering for free and pursuing healthcare-specific products
- Regulatory risk on the ad model — FDA/OIG/FTC scrutiny of pharma promotion adjacent to clinical recommendations could constrain the core monetization
- Liability from hallucinated or outdated clinical guidance — one high-profile patient-harm case could crater physician trust
- Journal partners raising license prices, demanding revenue share, or going direct with their own AI products
- Valuation risk — $12B against $100M+ disclosed / $150M estimated 2025 revenue (~80-120x) prices in flawless execution; any growth stall hits the mark hard
Moats, dependencies & bottlenecks
Moats
JAMA Network, Cochrane, NCCN, Wiley 400+ journals) Legally enforceable vs LLM scrapers, but rented not owned — renewal risk and journal counter-moves cap durability
Physician network / distribution (claimed >40% of US physicians daily, 10k+ facilities) Word-of-mouth standard-of-care status is powerful, but the product is free and switching costs are near zero absent EHR embedding
moderate-strong Doximity proved the model's stickiness with pharma buyers; premium CPMs persist only while engagement leads
Mount Sinai, Cedars-Sinai — Epic-integrated) high-if-achieved Workflow embedding is the classic healthcare lock-in; still early with three flagship systems, all on the free ad-supported model so far
medical-grade positioning) Asymmetric: slow to build, destroyed by one publicized harm event
Dependencies
NEJM Group (Massachusetts Medical Society) and JAMA Network (American Medical Association) content licenses Core answer-quality differentiator; renewal pricing power sits with the journals
Eli Lilly, Merck class of buyers) Single dominant revenue engine; cyclical and regulation-exposed; individual advertiser roster not disclosed
Nvidia ecosystem) Inference costs and model access shape gross margin; Nvidia is also an investor
The verified-prescriber gate is what makes the ad inventory premium
Health-system IT and EHR vendors (Epic ecosystem) for enterprise deployments Sutter and Mount Sinai deployments run inside Epic; Epic's own AI ambitions could help or squeeze the integration path
regulatory/reputational Hallucination liability and ad-conflict-of-interest scrutiny are existential tail risks
Advantages
- Category-defining brand ('ChatGPT for doctors') with dominant physician mindshare
- Near-zero customer acquisition cost via peer-to-peer physician adoption (65,000+ new verified clinicians/month)
- Premium licensed content (NEJM, JAMA, Cochrane, NCCN, Wiley) competitors have not replicated at scale
- Highest-value ad audience in media: verified prescribers at the moment of therapeutic decision
- Estimated ~90% gross margins and an asset-light cost structure
- ~$765M total raised, $450M of it since Oct 2025 — a fortress balance sheet vs seed-stage rivals
- Repeat-founder execution speed: $1B to $12B valuation in eleven months and ~7-9x usage growth in a year
Weaknesses
- Revenue concentration in one ad engine with regulatory overhang
- No audited financials; all key metrics are company-asserted or analyst-estimated
- Rented content moat with concentrated supplier power
- Free product means engagement, not contracts, is the retention mechanism
- Extreme valuation multiple leaves no room for execution stumbles
- US-only monetization today; international model unproven
Bottlenecks
- Ad-load ceiling: monetizing harder without degrading clinical trust limits near-term ARPU expansion
- Pharma ad-sales cycle — growing the advertiser base requires specialized, compliance-heavy sales teams that scale slower than usage
- Journal licensing scope — coverage gaps (specialty-society guidelines, non-licensed publishers) constrain answer completeness
- Enterprise procurement — hospital-wide EHR deployments face long security/legal review cycles despite bottoms-up demand
- Regulatory clarity — absence of settled rules for AI clinical-decision support plus point-of-care advertising caps how aggressively the model can be pushed
- US-centric verification: NPI gating does not translate directly abroad, slowing international expansion
Top signals & trends
Top signals
Top-tier crossover investors underwriting continued hypergrowth
The core leading indicator for ad inventory
Validates the enterprise second act and deepens switching costs; all currently on the ad-supported model
Private-market demand remains firm at-to-above the Series D mark; quotes are thin, model-derived, and not exchange prices
Widens the licensed-content moat beyond NEJM/JAMA/Cochrane
Incumbent response is underway with enterprise distribution OpenEvidence lacks
Reputational and regulatory kindling for the core revenue model
Commoditization pressure on the answering layer itself
Trends
OpenEvidence is the category leader in the fastest-adopting professional AI vertical
Structural tailwind for premium-CPM clinician platforms
Opens enterprise deals but favors incumbents (Epic, Microsoft/Nuance, Wolters Kluwer) in procurement
Rules for ad-adjacent clinical AI are unwritten; adverse rulemaking is the biggest model risk
Shifts value to proprietary content and distribution — OpenEvidence's bet — but compresses the pure-answering premium
Cheap capital now; mark-to-market risk if the AI multiple regime resets
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Multi-year license (Feb 2025) covering all NEJM Group content and multimedia from 1990 forward — cornerstone content supplier
Multi-year content agreement signed Jun 2025; official AI partner of JAMA and its specialty journals
Mar 2026 partnership covering 400+ medical journals and major references (Holland-Frei, Rook's, Yamada's)
GPU/compute ecosystem and strategic investor
Inference and hosting infrastructure; specific provider mix not disclosed
Free verified users — the audience product; company claims >40% of US physicians log in daily
Pfizer, Merck class) The paying side of the marketplace; individual advertiser roster not disclosed
Mount Sinai, Cedars-Sinai) Enterprise Epic-integrated deployments beginning Feb-May 2026, currently on the ad-supported model
The entrenched clinical-reference incumbent (decades of hospital contracts; Euronext-listed WKL with US ADR WTKWY); shipping genAI search across UpToDate
The proven pharma-ad-to-physicians platform (~$4B market cap, ~$645M TTM revenue, ~$228 ARPU); overlapping advertiser budgets and launching its own clinician AI tools
Massive owned medical corpus (The Lancet, Cell) plus ClinicalKey AI — a content-rich fast follower
Private; the free generalist many physicians also use — capability keeps improving and healthcare is a stated focus
Deep medical-AI research bench and Search distribution; GV is simultaneously an OpenEvidence investor
Owns the clinical-documentation channel inside EHRs and could bundle knowledge search
Private clinical-AI unicorn (raised at a ~$5.3B valuation, Jun 2025); ambient documentation today, adjacent to point-of-care knowledge tomorrow
Private evidence-search challengers competing on rigor and transparency; sub-scale but keep pricing/trust pressure on
Private EHR gatekeeper building native AI assistants; controls the workflow surface OpenEvidence embeds in