A free reference product makes it harder to justify a subscription merely for access to generated medical answers. Heidi publicly markets free Evidence access and CPD-related functionality, as checked on 22 September 2026. The resulting question for learning platforms is not whether they can also answer a question, but whether they help clinicians complete a different, worthwhile learning task. Heidi Evidence sets the competitive context.
This article is published by iatroX and includes iatroX among the tools compared. Referenced answers and CPD logging are not capabilities unique to iatroX, and we should not describe them as though they are.
What free evidence changes
A clinician seeking a cited explanation may reasonably begin with a free tool already available in their workflow. Heidi's Evidence page, checked on 22 September 2026, describes source-linked answers and source prioritisation alongside its wider clinical platform. The precise account and market entitlements still need checking. The Evidence product page is a description of the offer, not an independent test of every output.
This puts pressure on products whose main distinction is simply an additional answer box. A subscription needs to solve a problem that remains after the user can obtain an explanation. That might be guided revision, practice under constraints, feedback on reasoning or a coherent record of what the learner has actually done.
Those are hypotheses about value, not a claim that every clinician needs another subscription. For someone with an occasional reference question and no structured learning goal, free access may be sufficient.
Lower model costs and free access are different claims
Heidi's technical account dated 31 July 2026 describes fine-tuning an open-weight model for out-of-session Evidence. Its reported clinician-preference result concerns that task and evaluation, not the profitability of the free service. Heidi's model account explains the technical direction.
Operating a model more efficiently could make broad access easier to sustain, but free provision does not reveal the complete economics. Product development, source handling, support and distribution remain relevant. The September financing adds resources, yet it does not establish a profitable unit cost for each free answer. Heidi's 22 September 2026 financing announcement should not be read as such a calculation.
For competitors, the lesson is to understand the cost and purpose of their own free experience. It may be valuable in its own right, a route into another useful product, or both. None of those interpretations requires disparaging a competitor's free offer.
Reference is an input to learning, not always the whole task
Consider a fictional GP preparing for an assessment who repeatedly confuses two decision pathways. Reading a clear answer can resolve today's uncertainty, but the learner still needs to explain the distinction without looking at the answer next time.
A useful practice sequence would ask for an initial commitment, explore the reasoning, expose the misconception and revisit the distinction in a different scenario. The important comparison is whether the learner performs that sequence, not whether the interface contains a long explanation labelled "education".
Guided questioning is not exclusive to specialist medical products. The case for a specialist learning subscription must therefore rest on how questions, feedback, progression and the relevant professional goal fit together. It cannot rest on an inaccurate claim that general-purpose AI only supplies instant answers.
What iatroX adds, and what remains free
Per iatroX product information, September 2026, Ask-iatroX and free question access are genuinely free, with no trial expiry or verification gate. Ask-iatroX draws on NICE, CKS, SIGN and SmPC information from emc with linked sources. Its published methodology describes retrieval, ranking, grounding, output checking and uncertainty handling; these are design features, not a guarantee of correctness. Ask-iatroX and the methodology describe that reference layer.
The paid learning proposition, per iatroX product information in September 2026, combines question banks covering more than forty examinations, adaptive sequencing, spaced repetition, Socratic Tutor, the study planner, iatroX Simulations and CPD tools. The Tutor opens on an attempted question and asks targeted follow-ups rather than simply supplying the answer. The question-bank entry point and Tutor are the relevant product routes.
The September 2026 PARA bank, formerly PANE, is a distinct free offer with more than 1,000 questions, adaptive sequencing, spaced repetition and a performance dashboard; ongoing Socratic tutoring is paid. PARA concerns the UK physician associate registration assessment, not the US PANCE or PANRE. The existing PANE-path page remains the verified destination despite the assessment-name change.
One relevant goal, several methods
For the fictional GP, value might come from using a question to reveal a gap, a Tutor exchange to explain it and a later case to test transfer. Per iatroX product information, September 2026, the study planner uses the examination date and daily study-time input, adapts to quiz performance and organises work through foundation, application and performance phases. These are planning features, not a prediction of an examination result. iatroX's learning entry point provides the product context.
The commercial argument is not that the GP can access unrelated specialty examinations. It is that different learning methods can support the one goal that matters to that learner. A larger catalogue is irrelevant when the necessary content, feedback or workflow is missing.
At the prices supplied by iatroX for September 2026, the complete paid bundle costs £99 a year, paid upfront, equivalent to £8.25 a month billed annually, or £29 a month. Simulations and CPD tools are included, not separate add-ons. Three monthly payments total £87 and four total £116, so the annual option becomes cheaper from the fourth month; a shorter period may favour monthly access. These are prices for the learning bundle, not charges for free Ask-iatroX.
CPD should describe learning, not merely activity
Heidi markets CPD-related functionality as of 22 September 2026. iatroX should therefore explain its own workflow rather than imply that competitors cannot record learning. Heidi Evidence makes that boundary clear.
Per iatroX product information, September 2026, the learner selects profession and practice context, reviews and personalises the draft CPD record, and can export a PDF or send it directly to FourteenFish through a linked account. Completed evidence remains accessible after the subscription ends. These are record-management features, not a universal claim of formally accredited CME. iatroX CPD is the relevant destination.
The existing Heidi Evidence versus iatroX comparison covers the product-level scenarios. This article is its funding-and-strategy companion, concerned with what a free reference offer changes about the case for structured learning.
Verdict by learner scenario
For rapid cited reference within an existing Heidi workflow, free Evidence is a credible option to examine. For a UK-facing guideline question without a study requirement, free Ask-iatroX may meet the need. For structured revision, assess the complete practice-and-feedback journey and the duration for which paid access is useful. For CPD, judge the quality of the personally reviewed record rather than treating logging alone as proof of learning.
Frequently asked questions
Is referenced clinical information unique to iatroX?
No: Heidi also publicly describes source-linked Evidence answers as of 22 September 2026. The useful comparison concerns source relevance, the task and the surrounding workflow.
Does iatroX charge for Ask-iatroX or basic free question access?
No: per iatroX product information, September 2026, those remain free without trial expiry or a verification gate. The subscription covers the combined advanced learning tools.
Does a CPD export automatically provide accredited CME?
No: a professional learning record and formal accreditation are different claims. The learner should check the applicable requirements and personalise the record to reflect actual learning.
Explore iatroX's combined question-bank and learning experience →