OpenEvidence’s 2026 announcements describe a platform expanding in several directions at once. Clinical investigation, professional education, institutional deployment and the practical work of obtaining treatment now sit within the same broader product story.
The latest development is Access Intelligence, announced on 11 November. Understanding its significance requires looking beyond that individual launch. A deeper research option, a learning activity and an administrative action can all begin with a clinical question, but they ask different things of the system and its user.
This guide maps selected material announcements and current documentation as checked on 11 November 2026. It distinguishes released functions, research previews, content partnerships and evidence of performance. It is a product analysis, with no claim to have tested account-restricted functions or measured comparative clinical outcomes.
For iatroX, the expansion makes the competitive task clearer. The positive case for a specialist platform rests on the reference and learning work it helps clinicians complete, with an intelligible path from a question to a better understanding.
A dated map of the main developments
| Date or period | Development | What changed in the published offer |
|---|---|---|
| March 2026 | Coding Intelligence | Visits documentation gained coding suggestions and supporting rationale. |
| 28 July 2026 | Education platform | Accredited continuing education and selected certification activities were announced. |
| 3 September 2026 | Model family | Osler, Sackett and Snow production choices; Darwin research preview. |
| 11 September 2026 | Canadian access | A MINC verification route for Canadian physicians. |
| Late October 2026 | NYU Langone collaboration | Patient-aware evidence within an Epic deployment was announced. |
| 4 November 2026 | Learn | Question practice, simulated cases and flashcards joined the learning offer. |
| 6 November 2026 | ACR patient education | English- and Spanish-language fact sheets entered the content network. |
| November 2026 | ACP content agreement | Clinical guidance and Annals of Internal Medicine content were licensed. |
| 11 November 2026 | Access Intelligence | Affordability and access support extended the post-decision workflow. |
Sources and the relevant limits are discussed below. The table is a map of selected product milestones, not a claim to catalogue every partnership or company announcement during the year.
The chronology shows why an older description can become incomplete without becoming wholly wrong. Evidence investigation remains important, while the surrounding jobs have multiplied.
March brought documentation closer to reimbursement
The March Coding Intelligence announcement describes ICD-10, E/M and CPT suggestions derived from Visits documentation, including medical-decision-making rationale. OpenEvidence’s company-issued announcement
The strategic significance is the connection between the encounter and the work generated by it. A note is a clinical record, but its content can also affect administrative processes. Bringing those processes together changes what needs review.
A useful evaluation should therefore distinguish fidelity to the encounter from the suitability of a suggested code. A complete-looking document can still contain an unsupported statement. A coding suggestion can require a different check from the clinical explanation beside it.
The corresponding opportunity for specialist knowledge products is precision about their role. Helping a clinician understand the source behind a decision is valuable even when another product prepares the documentation or reimbursement material.
July connected clinical questions with accredited education
The 28 July release describes an AKH-accredited activity linking searched evidence, reflection and participating certification requirements. Eligibility and credit arrangements are specific, including NPI-verified professions and selected boards. OpenEvidence’s education announcement
This milestone matters because reading, learning and earning credit are related activities with different requirements. A useful explanation can improve understanding without qualifying for a particular credit. A credit-bearing activity still needs to be relevant to the clinician’s development.
The comparison with iatroX should preserve that distinction. Its CPD workflow produces learning records the user reviews and confirms, with supported exports. Those records should not be described as interchangeable with every accredited CME or certification arrangement.
A clinician choosing between activities should start with the purpose: a knowledge gap, a professional requirement or both. The quality of the learning and the acceptability of the record each deserve a clear answer.
September made investigation depth a product choice
The model-family release positions Osler as the default, Sackett as deeper interactive investigation and Snow as extended production research succeeding DeepConsult. Darwin is a separate application-only research preview. OpenEvidence’s 3 September model announcement
That structure gives the user a decision before the answer: how much investigation does this question require, and how much time is available? The appropriate choice depends on the task rather than the prestige of a model name.
A bounded source check and a disputed specialist question can legitimately need different treatment. Longer investigation may uncover more relevant material, but time spent generating an answer does not itself establish its correctness.
The release also reports benchmark results for Darwin. Those results concern a particular evaluated configuration; they do not establish that ordinary production accounts have the same capabilities or that the system performs equally across every clinical task.
A reader should therefore keep four identities distinct: the name shown in the product, the model or system evaluated, the available tools and the deployed workflow. A score detached from those details can tell the wrong story about what a clinician actually receives.
Our earlier model-selection guide covers the individual choices. The important 2026 change is that investigation depth became an explicit part of the user’s workflow.
International access expanded through particular routes
The 11 September MINC announcement supports free access for verified Canadian physicians. It is a positive, defined eligibility route rather than a complete international availability list. OpenEvidence and MINC’s announcement
That distinction is useful for international readers. A platform can expand across borders while retaining restrictions in others. A verification route for one profession does not establish an identical route for every trainee, clinician or organisation.
Access and applicability also need separate consideration. Being eligible to use a service does not establish that an answer follows the relevant local pathway or that a learning activity matches the next examination.
For iatroX, international educational relevance is a positive competitive opportunity. A clinician can choose its supported preparation pathway because of the qualification they are pursuing, even in a market where another evidence platform is available.
Institutional distribution changes the evidence experience
OpenEvidence’s current Enterprise page describes SSO, EHR launch and patient context where supported, together with institutional agreements and data controls. It is offered at no cost to eligible provider organisations supporting verified US clinicians. OpenEvidence Enterprise
The announced NYU Langone collaboration provides a concrete Epic-related example of this direction. Its stated scope is patient-aware evidence inside the organisation’s workflow. NYU Langone and OpenEvidence’s announcement
An institutional deployment changes the source of context, the route into the tool and the responsibilities surrounding its use. Those details should be examined independently of the answer’s clinical content.
For example, a reviewer needs to know whether a statement came from the patient record, a clinical reference or a generated interpretation. A context-rich response can still miss relevant information or carry forward an unresolved conflict.
Institutional integration also does not make a separate learning routine redundant. A trainee preparing for an assessment may want a deliberately chosen practice session rather than an answer appearing during record review.
November made the learning overlap unmistakable
Learn’s 4 November launch adds more than 50,000 board-style questions, Preceptor simulated-patient tutoring and generated flashcards alongside CME. It is free to OpenEvidence users, subject to actual account eligibility. The Learn announcement
That makes the educational comparison more demanding. A specialist platform cannot establish its value merely by listing questions, cases and AI discussion. It needs to show how those activities fit the learner’s particular objective.
Relevant questions include where the session begins, what context the tutor has, how feedback addresses an attempted answer and what guides the next activity. A learner preparing for an examination also needs to understand the coverage and structure of the selected material.
Our updated Learn, AMBOSS and iatroX comparison follows those tasks directly. The central issue is the learning sequence, rather than whether a supplier began as a reference company or an education company.
Content agreements expand the available evidence
The ACR’s 6 November release adds patient fact sheets in English and Spanish, surfaced with source links. This is a patient-education development rather than a new diagnostic model. The ACR’s announcement
ACP’s November announcement adds guidance and Annals of Internal Medicine material. Its own page carries a 5 November dateline, while its news listing and issued release are dated 9 November. ACP’s page, ACP’s issued release
A licensed source can improve the material available for retrieval. It does not establish which passage supports a particular answer or whether the synthesis preserves the important qualifications.
The clinician still needs to distinguish the original recommendation from the platform’s interpretation. That is especially important when several interfaces draw on the same source: their agreement may reflect a shared starting point rather than independent confirmation.
Access Intelligence extends the post-decision workflow
The 11 November Access Intelligence release brings support-programme discovery, enrolment and prior-authorisation drafting into the platform’s access offer. The launch announcement
Its significance is the move from understanding a treatment choice towards dealing with the practical work that can follow it. Each step needs its own definition of completion.
Finding a relevant programme, preparing a form and obtaining a decision are different achievements. A service evaluating the workflow should be able to identify the status of each case and the responsible next person or system.
The dedicated Access Intelligence analysis explains the released functions and the evidence needed to assess operational benefit. Document generation alone does not establish that care became more affordable or accessible.
Follow information as it moves between tasks
Consider a fictional clinician who investigates a question, prepares a document and then revisits the topic for learning. The same clinical concept appears at each stage, but the required output changes.
The investigation should leave a supported explanation and visible uncertainty. The document should preserve the facts and qualifications relevant to its recipient. The learning session should give the clinician an opportunity to work on what they did not understand.
Moving between those tasks is convenient only when the information retains its meaning. A qualification should not disappear because a longer answer became a shorter letter. A learning activity should not silently treat an unresolved clinical inference as an established fact.
That creates a useful test for any connected platform: can the user identify where a statement originated, what changed during the transition and what still requires their judgement? The number of functions matters less than the reliability of the handoff between them.
A specialist reference-and-learning routine can complement a wider workflow precisely because it creates room to examine that underlying distinction before carrying it into another task.
Released, preview and announced are not synonyms
The product map can be read through three questions. Is the function available to the intended account? Is the supporting connection or content live in that configuration? What evidence exists about its performance?
A production function can have limited geography or institutional requirements. A research preview can produce interesting results while remaining unavailable for routine use. A partnership can establish an intended relationship without providing a complete implementation timetable.
These distinctions should remain visible in buying and evaluation decisions. They prevent the capabilities of a restricted preview, an individual account and an institutional deployment from being merged into an imaginary universal product.
OpenEvidence’s current standard terms, updated on 17 September, exclude UK, EU and EEA access. Particular properties and applicable institutional agreements can differ. Its broader international activity should be understood alongside those conditions, rather than treated as evidence of unrestricted global availability.
Where iatroX fits after these announcements
iatroX’s case is the connection between a reference question and deliberate professional development. Its FAQ describes guidance-linked reference alongside question practice and an attempted-question Tutor, with educational context kept explicit.
That gives the clinician a purposeful place to work on the distinction behind an answer. A wrong choice can become a focused discussion; a recurring uncertainty can become a practice objective; a useful lesson can inform the next session.
The value is practical even when another system supplies an evidence report, prepares documentation or supports an access process. Clinicians still need to understand what they rely on and develop the judgement they bring to the next case.
Ask and selected banks are free; wider learning uses the current all-access offer, £29/month or £99/year in the UK; US$29/month or US$99/year elsewhere. Choose the activity needed, rather than assume every professional task should be consolidated into the same service.
OpenEvidence’s 2026 expansion is substantial. It also makes specific product purposes more important. iatroX can compete confidently on relevant reference, examination context and a connected learning routine, while the broader market continues to bring evidence, education and administrative work closer together.