DR.INFO’s current product is worth looking at afresh. Its October 2026 instructions cover both the web service and an iOS app, while its reference workflow now explicitly describes country-specific guidelines and questions anchored to individual document passages.
Those details matter more than a broad claim to be an “AI for doctors”. They tell a clinician what to open, how to examine a source and which task the product is intended to support.
This overview looks at the public offering as checked on 30 October 2026. It builds on our earlier DR.INFO review, focusing on current access, source handling and the evidence behind the product. It also explains where iatroX’s question banks, Tutor and CPD workflow serve a different learning need.
The current reference workflow
DR.INFO is built by Synduct GmbH in Munich. Its instructions for use, version 1.2 dated 13 October 2026, describe four core activities: answering medical information questions, searching drug information, finding clinical guidelines and generating visual abstracts.
These activities offer different ways into the same underlying problem: finding the relevant part of a large medical document.
A useful evaluation should therefore include more than one general question. Try an information question requiring two sources, a search for a named guideline, and a product-information lookup. Then examine whether each route makes the original passage easy to reach.
The iOS App Store listing describes an additional practical feature: saving answers, medicine summaries and visual abstracts in a personal library, alongside conversation history. For someone returning to an unfamiliar topic between teaching sessions, that continuity may be more useful than generating another answer from scratch.
The October instructions apply to the web application and the iOS app from version 1.0.10. App access is therefore a concrete part of the current offering, rather than a future mobile promise.
Country settings and passages deserve attention
The current FAQ says national guidelines depend on the user’s country setting. It specifically describes German-language retrieval, including AWMF and DGVS sources, while international and European society guidance remains available. It also acknowledges that coverage of any individual organisation is incomplete.
That is a more useful description than assuming “European sources” means a complete national reference library.
For a German clinician, a practical test is to start with a named German guideline and check the retrieved version against its official register entry. For a UK clinician, the question is which UK source actually supports the answer. A country preference should help find material; the cited document establishes what was used.
The FAQ also describes passage-level interrogation of guidelines, SmPCs and consensus statements, with citations anchored to the relevant text. This can help when a broad summary conceals an important condition.
An educational question might be: “Which population does this recommendation concern, and where does the document define it?” The value lies in connecting the explanation to the source’s scope, rather than collecting another confident paraphrase.
Separate reference access from the documentation roadmap
The homepage labels scribe and discharge-letter functionality “Coming Soon”. Meanwhile, the pricing page lists documentation and record-system connections under organisation arrangements.
For an individual user, the concrete starting point is the reference product. The published free tier includes 30 clinical questions and five visual abstracts per month; paid individual plans advertise unlimited use subject to applicable limits.
A practice evaluating documentation should request a demonstration of the exact proposed package and its deployment requirements. A listed organisation feature and a generally available feature in an individual account are different purchasing propositions.
This distinction keeps a product trial focused: first assess the activities you can actually use, then examine any additional workflow being offered to your organisation.
What the current label means for the questions you ask
The September 2026 product label describes clinical reference and knowledge support. It also excludes analysing personal patient health data, calculating doses and providing diagnostic or therapeutic recommendations. Its precautions prohibit patient details even when identifying information has been removed.
For a useful trial, frame general professional information questions or fictional teaching problems. A question about the evidence underlying a guideline recommendation fits that reference task. Uploading a patient narrative to obtain a treatment choice changes the task substantially.
This is especially relevant when comparing reference search with the broader documentation marketing. The precise service, intended use and data-handling arrangement need to match the work being proposed.
The label is also a reason to distinguish a retrieved dosing section from a dose calculation. Reading the source can help explain a rule; it does not establish that the software has calculated an individual regimen.
Two kinds of evidence, two different questions
DR.INFO has published evidence about both benchmark performance and perceived usefulness. They should be read separately.
The HealthBench preprint’s third version, revised on 27 July 2026, reports a score of 0.68 on the 1,000-conversation Hard subset. It also describes a separate 100-conversation comparison with other clinical assistants. The authors disclose employment with Synduct, which supported the work.
That version matters: figures from an earlier abstract or product page should not be silently combined with the revised evaluation. A benchmark score is also not a percentage of patients receiving correct treatment. It reflects performance against the evaluation’s scoring framework.
For a clinician, the useful follow-up questions concern the tested system version, comparison dates, scoring method and similarity to the information tasks you actually perform. An evaluated text-answering task does not establish performance across every advertised feature.
The prospective pilot report asks a different question. Twenty-nine physicians and medical students in Portuguese institutions used DR.INFO v1.0 for five working days during an October 2025 study window. Sixteen completed the final evaluation. The report describes positive perceptions, while explicitly calling for objective accuracy verification and larger controlled studies.
That gives information about early user experience. It should not be read as evidence that today’s app improves patient outcomes. Our pilot-study analysis examines those findings in more detail.
For a broader shortlist, our OpenEvidence alternatives guide compares reference, documentation and learning workflows.
Where iatroX offers a different learning workflow
DR.INFO’s source-reading features address finding and explaining information. A learner also needs to discover which distinctions they can recognise without the source already open.
On iatroX, the exam catalogue connects revision with an available assessment pathway. Tutor works from an attempted question, retaining its context while exploring the learner’s reasoning. That makes a wrong answer a starting point for teaching.
For UK product-information questions, iatroX’s Datapharm partnership brings emc material into the clinical reference workflow. Country still matters: a UK product document and a German guideline answer different questions about the setting in which information applies.
The practical comparison is therefore about the activity. Someone preparing a teaching summary may value DR.INFO’s document interrogation and visual abstract. Someone trying to correct a recurrent examination error may need a question, an explanation and a later opportunity to retrieve the distinction independently.
A concrete way to connect search and practice
Consider a fictional revision explanation that quotes a recommendation for adults, while the question concerns a different population. The quotation is genuine; its applicability is the problem.
First, identify the mismatch before asking for another summary. Open the source and locate the population definition, exclusions and surrounding conditions. Write a short explanation of why the quoted passage does not settle the question.
Next, attempt a different question in which the population is appropriate but another condition changes. This checks whether you learned to inspect scope, rather than memorised the first correction.
A source-reading tool can help with the first stage. Question practice and discussion can expose whether the principle transfers to the second. Keeping those activities distinct makes the learning easier to evaluate.
For a professional record, iatroX CPD offers focused question sessions, Tutor-supported review and a private reflection draft that the clinician personalises and confirms. The useful entry describes the mistaken assumption, the source checked and the change in reasoning. Reading a product review becomes valuable when it leads to an observable learning activity.
Frequently asked questions
Can I use DR.INFO on an iPhone?
A native iOS app is listed in the App Store, and the current instructions cover it alongside the web application. Check device requirements and the features included in your chosen plan.
Does DR.INFO’s benchmark establish that it is the best tool for my work?
The reported evaluation addresses defined text tasks and system versions. Use it to examine the methods and tested behaviours, then assess the sources and workflow relevant to your own professional information needs.
When does iatroX become useful after a referenced answer?
When the answer reveals a distinction you need to practise. An attempted question and Tutor discussion can expose the reasoning gap; a focused CPD session can help you record the learning and plan a follow-up.
Turn a source question into a learning session
Use iatroX CPD to practise one distinction through questions and Tutor, then review and finalise a record of what you learned.