Tandem and Tortus are two of the clinical AI tools most visible across UK general practice. Both reach large footprints through distribution partners — Tandem through Accurx, which touches the overwhelming majority of GP practices, and Tortus through X-on Health's Surgery Intellect, deployed across thousands of practices. Both are ambient tools that transcribe, summarise, draft letters and suggest codes. Where they differ is in the regulatory class each has achieved and in how each presents its evidence.
Regulatory posture
This is the clearest distinction. Tandem's AI scribe and coding assistant are CE marked under EU MDR as Class IIa medical devices, externally assessed by a notified body — a deliberate move to the higher of the two relevant classes for this kind of software.
Tortus is MHRA-registered as a Class I medical device, with Class IIa reported as pending. In other words, Tortus is on the same ladder and climbing, but Tandem currently holds the higher achieved class. Both are compliant with the expected NHS governance frameworks — DTAC, DSPT, GDPR and, in Tortus's case, Cyber Essentials Plus.
For a buyer, the practical reading is that Tandem can point to a completed higher-class assessment today, while Tortus is positioned to reach the same rung but has not yet, on current reporting.
Clinical evidence
Interestingly, the two tell their evidence stories from opposite ends.
Tortus leans on published performance data: independent audits reportedly showing a 97% clinical accuracy rate, evaluations indicating 35–40% efficiency gains per clinical session, and an average of around four minutes saved per consultation. That is a comparatively concrete, operational evidence base for a scribe.
Tandem's emphasis is more regulatory than study-led — its primary assurance is the notified-body Class IIa assessment of its scribe and coding products, rather than a set of published accuracy figures. Neither approach is inherently superior: one foregrounds independent regulatory assessment, the other foregrounds measured real-world performance. A careful buyer would want both, and should ask each vendor for the evidence it leans on least.
Scope and direction
Both are widening beyond pure transcription. Tandem has extended into coding and, most recently, in-consultation clinical decision support, all under Class IIa. Tortus generates structured summaries, referral letters and suggested clinical codes in real time, with its higher-class assessment pending. The trajectories are similar; the timing differs.
| Dimension | Tandem | Tortus |
|---|---|---|
| Core function | Scribe + coding + in-consultation CDS | Ambient scribe + coding suggestions |
| Regulatory class | CE marked EU MDR Class IIa (notified-body assessed) | MHRA Class I; Class IIa reported pending |
| Distribution | Accurx (reaches ~98% of GP practices) | X-on Health "Surgery Intellect" (3,500+ practices) |
| Headline evidence | Class IIa regulatory assessment | ~97% accuracy audits; 35–40% efficiency gains |
| Governance | ISO 13485/27001, NHS DSPT, GDPR | DTAC, DSPT, GDPR, Cyber Essentials Plus |
| Record-system integration | EMIS, SystmOne + 100+ systems | Consultation-workflow integrated (via X-on) |
Where iatroX sits
Tandem and Tortus are both documentation-and-workflow tools — they capture the encounter, code it, and increasingly support decisions within it. iatroX occupies a different layer entirely: it is the clinical knowledge and learning layer, not a scribe.
That makes iatroX complementary rather than competitive with either. A clinician can run whichever scribe their practice has chosen and still use iatroX — a free, UK-registered (UKCA-marked Class I) clinical AI grounded in NICE, CKS, SIGN and the SmPC — for fast, cited clinical answers and for exam and CPD learning. iatroX does not claim a higher regulatory class than Tandem; its role is to be the trusted UK knowledge layer alongside the documentation tools, not to replace them. And while its Class I registration is self-declared rather than notified-body assessed, it is not ungoverned: it is backed by a full clinical risk-management file and safety case, independently reviewed.
Frequently asked questions
Which has the higher regulatory class, Tandem or Tortus? Tandem currently holds the higher achieved class: its scribe and coding assistant are CE marked under EU MDR as Class IIa devices. Tortus is MHRA-registered as Class I, with Class IIa reported as pending.
Does a higher regulatory class mean a more accurate scribe? Not directly. Regulatory class reflects independent assessment of intended use, risk management and quality systems — not a guarantee of accuracy on any given note. Tortus, at Class I, points to published accuracy and efficiency figures, illustrating that regulatory class and measured performance are separate questions.
Are Tandem and Tortus used at scale in the NHS? Yes. Tandem reaches NHS clinicians through Accurx, which touches the large majority of GP practices, and Tortus is deployed across thousands of practices via X-on Health's Surgery Intellect platform.
How does iatroX relate to these scribes? iatroX is a knowledge and reference layer, not a scribe, so it is complementary. Clinicians can use it alongside any documentation tool for cited UK clinical answers and for learning, as a free, UK-registered Class I clinical AI grounded in NICE, CKS, SIGN and the SmPC.