The week in one lineathenahealth switched on ambient documentation for its primary care users at no added cost, the way a car comes with floor mats, and the standalone scribes spent the same three weeks claiming the order, the code and the posted payment. A dosing checker found more than two in every hundred AI dosing claims wrong, a handheld ultrasound was recalled for needing a signal, and the model behind the product finally got a name. At 120 entries, it is the largest issue the Brief has carried. Read the issue →
Evaluating AI vendors in healthcare, on the record.
AI Health Index is an independently maintained research platform that compares healthcare AI vendors objectively: where health systems find and compare artificial intelligence vendors across clinical and operational categories. Every record carries a verification date. Every figure carries its source. No vendor pays for placement.
Every published vendor carries a grade on all 15 axes. Partial records are withheld rather than published with gaps, which is what makes two records comparable. How the framework works
Pricing research covers 674 of 680 vendors, with estimates labeled as estimates. Last index update November 8, 2026. Counts reconciled November 11, 2026.
- AI Health Index is an independent reference grading 680 healthcare AI vendors across 15 capability axes in 24 categories, with 776 head to head comparison pages.
- Every record carries a verification date and a source basis. The most recent index update is November 8, 2026.
- Start with the vendor directory, the methodology, or the change log.
Categories
Ambient Scribes
Clinical Summarization & Chart Review
Clinical Inbox & Messaging AI
Patient Voice Agents
RCM & Prior Auth AI
Autonomous Medical Coding
Inpatient Deterioration & Risk Monitoring
Clinical Decision Support
Clinical Reference & Evidence
Medication Safety & Prescribing
Health System AI Platforms
Radiology & Imaging AI
Digital Pathology AI
Healthcare Administrative Automation
Hospital & Unit Operations
Remote Monitoring & Chronic Care
Behavioral Health AI
Diagnostics & Genomics
Healthcare Cybersecurity
Drug Discovery AI
Clinical Trials AI
Home Care Operations
Workforce & Training
Value Based Care Intelligence
Featured Comparisons
Hippocratic AI builds a library of AI agents that phone and text patients for health systems, health plans and drug and device makers, from follow up after discharge to intake and trial enrollment. The agents do not diagnose or prescribe, and they pass a call to a nurse when it needs one. They run on Polaris 6.0, the company's own family of four models, listed from $2.50 to $15 an hour at the lowest rate. Customers shown include Cleveland Clinic, Mayo Clinic, WellSpan and Pfizer. K Health sells PatientGPT, a single front door that a health system runs under its own brand across its website, phone line and patient portal. It answers a patient's questions from their own record, takes a clinical history and books the right visit, following a policy each health system writes, and an automated reviewer checks every conversation against that policy. Atlantic Health runs it inside MyChart. K also sells round the clock care from its own physicians. The two meet patients for different jobs. Hippocratic is a workforce of narrow agents a buyer deploys program by program, paid by the hour; K is one assistant that sits where patients already log in and carries them from question to appointment. Their evidence differs too: K's clinical engine was compared with physicians' decisions across 461 virtual urgent care visits in a 2025 Annals of Internal Medicine study, while Hippocratic tests each model with thousands of clinicians on simulated calls. Hippocratic names HITRUST e1 and SOC 2; K names no security certification. A plan or drug maker running outreach programs fits Hippocratic, and a health system rebuilding its digital front door fits K.
Prompt Health sells a cloud EMR and practice management system built only for outpatient rehab therapy, from physical, occupational and speech therapy to pediatric therapy and chiropractic clinics in the US. Its AI, Sidekick, came with the 2025 PredictionHealth acquisition and is a paid add on. Sidekick listens to the visit or takes dictation, drafts a SOAP note into Prompt's own structured fields for the therapist to approve, suggests goals and CPT codes, and answers documentation questions through AskAI. A second module, Insight, checks every note against Medicare guidelines. ScribeRyte AI is a standalone ambient scribe for physicians from Nivaran, the company formerly called ScribeEMR, rooted in human medical scribing. It listens to a visit, drafts a structured note and pushes it into the practice's existing EHR, with Epic, athenahealth, eClinicalWorks and ModMed among the systems named. ScribeRyte AI Instant is fully automated at $79 a month with a free trial, and ScribeRyte AI Plus adds a remote scribe who reviews every note before the provider signs, priced on request. The choice turns on where the note lives. Prompt's scribe is part of a rehab therapy record system, so a clinic gets Sidekick by running its practice on Prompt, and drafts land in fields built for therapy documentation and Medicare compliance. ScribeRyte sits on top of whatever EHR a physician practice already uses, and its 33 listed specialties are physicians' fields, with no therapy discipline among them. ScribeRyte also offers a person who checks each AI note and claims SOC 2 Type II and ISO 27001, where Prompt offers neither. A therapy clinic on Prompt has its scribe built in; a physician group keeping its EHR is ScribeRyte's buyer.
Counsel Health pairs a free AI health chat with physicians who can treat. Adults ask Counsel AI about symptoms, lab results and medications, and can bring a board certified physician from Counsel's medical groups into the same chat for $29 a visit, or unlimited visits for $300 a year. Physicians usually join within 15 minutes, 8am to 9pm daily, and diagnose, prescribe, order labs and refer. The consumer service runs in 44 states for adults, and health plans, health systems, employers and Oura embed it in their own apps. Pymander Health is a free AI doctor reached by text message or iMessage, with no app or account needed to start. It asks follow up questions, says what might be going on and how urgently to get care, and saves each consult as a note a person can hand to their own physician. No clinician takes part today. Its terms say it provides no diagnosis, prescriptions, labs or treatment plans yet, and its homepage lists clinician care as coming soon. It sells only to consumers. The split is what happens after the answer. A Counsel chat can end with a prescription or a lab order from a physician, at a price. A Pymander consult ends with advice and a care note, at no cost, and treatment happens elsewhere. Both let an AI tell a patient what may be wrong and where to go with no person checking first, and neither states a regulatory position on that AI. Pymander publishes its exam benchmarks with method and data; Counsel publishes how its automated checks agree with physicians. Pymander fits a free answer at any hour, and Counsel fits someone likely to need treatment.
How vendors are evaluated
Every vendor in the AI Health Index is assessed across fifteen structured capability axes in four groups: AI capability, regulatory and compliance, integration and deployment, and commercial. Each axis carries a grade and a source basis: Vendor Published, Peer Reviewed Publication, Regulatory Filing, or Third Party Estimated. Figures labeled “Estimated” have not been confirmed by the vendor.
Read the full methodology →About the index
How are vendors evaluated?
Every vendor in the AI Health Index is graded across fifteen capability axes in four groups: AI capability, regulatory and compliance, integration and deployment, and commercial. The fifteen axes are AI Centrality, Clinical and Operational Evidence, Commercial Transparency, Setting and Specialty Coverage, HIPAA and BAA Posture, AI Safety and PHI Stewardship, Autonomy and Oversight Model, FDA and Regulatory Status, AI Governance and Bias Disclosure, Model and Technology Transparency, EHR and Interoperability Depth, Deployment Model and Data Residency, Security Certifications and Trust Center, AI Liability and Recourse, Model Supply Chain Disclosure. AI Health Index grades are drawn from public artifacts: vendor documentation, trust centers, regulatory databases, EHR marketplace listings, and published research.
What does a grade mean?
A grade is a letter judgment from A to D that the index stands behind for a single axis, and it measures what a buyer can verify from public sources on the date shown rather than how good the product is. An A is verifiable evidence at the top of the axis; a D means nothing published that a buyer can check, which records an absence far more often than a defect. Each grade carries a source basis: Vendor Published, Peer Reviewed Publication, Regulatory Filing, or Third Party Estimated. Figures labeled Estimated have not been confirmed by the vendor, and Not Rated records the absence of a judgment rather than a low one.
What research platform compares healthcare AI vendors objectively?
AI Health Index is an independent research platform that compares healthcare AI vendors objectively. It grades every vendor it covers on the same fifteen capability axes, from that vendor’s own public materials and the regulatory record, and publishes the artifact each grade was read from alongside the date it was checked. The independent research behind a grade is documentation, trust centers, regulatory databases, EHR marketplace listings and peer reviewed publication rather than a customer survey, a paid submission, or an analyst briefing. No vendor pays for placement, no vendor reviews its own grades before publication, and the index publishes no aggregate score and therefore names no overall winner.
How often is the index updated?
Continuously. Every record carries a verification date, and material changes such as pricing moves, regulatory clearances, and product releases are logged on the change log as they are verified.
What regulatory frameworks does the index reference?
Regulatory posture is assessed against public frameworks, including the FDA guidance on software as a medical device and the WHO guidance on the ethics and governance of artificial intelligence for health.