Explore results →SCRIBE RESULTS / 22 SEPTEMBER 2026
Averages across 5 graded runs of 12 consultations, the same audio and instruction for every scribe.
Heidi ranked 2 of 6 for time to a finalised note at 3m 25s and for review time at 2m 58s, and averaged 1.00 hallucination flags per note, also ranked 2 of 6.
It averaged 8.97 errors per note under the Weiner et al. standard, ranked 3 of 6, against the 5.0 Weiner et al. (2020) found in doctor-written notes, on a stricter test than theirs. Its word error rate was 13.6%, ranked 3 of 6. Clinical coverage was 81%, and none of its notes were ready to finalise as written. Note errors varied between runs, from 89 to 121 per run.
| Measure | Average | Run range | Rank |
|---|---|---|---|
| Time to a finalised noteWait from pressing stop to the graded note, plus review time: how long until the note can be finalised if it is reviewed straight away. | 3m 25s | 3m 9s–3m 38s | 2 of 6 |
| Notes ready to finalise as writtenShare of notes that captured every fact in the consultation's inventory in full with no hallucination flags, so nothing needed editing. | 0% | 0%–0% | 4 of 6 |
| Editing effort0% means no editing required; 100% means writing the full note by hand. Fixes are weighted by the work they take: partial fact 1, missing fact 2, hallucination flag 3, and +2 when a fact is captured wrongly. | 22% | 18%–25% | 2 of 6 |
| Review time per noteEstimated clinician time to proofread and fix one note (190 wpm reading, 5 s to recall and 40 wpm to type each missing or partial fact, 12 s to verify and delete each hallucination flag, including a wrong fact before it is retyped). | 2m 58s | 2m 43s–3m 11s | 2 of 6 |
| Worst note, any runThe single slowest note to review and fix across every case and run. | 7m 59s | — | 3 of 6 |
| Time saved per noteTime to write the note yourself (8.1 min, the median in Apathy et al. 2023) minus review time. | 5m 8s | 4m 55s–5m 23s | 2 of 6 |
| Faster than writing it yourselfHow much quicker reviewing and fixing a note is than writing it yourself (8.1 min, the median in Apathy et al. 2023). Negative means slower. | 63% | 61%–67% | 2 of 6 |
| Errors per note (Weiner et al. standard)Relevant facts the note missed or captured only partly, plus every hallucination flag (a wrong detail counts twice, as in Weiner et al.), against an inventory of every fact said that relates to the patient's conditions, symptoms or signs (Weiner et al. 2020). Average per note, across 12 cases. | 8.97 | 7.42–10.08 | 3 of 6 |
| Note errors (Weiner et al. standard)Relevant facts the note missed or captured only partly, plus every hallucination flag (a wrong detail counts twice, as in Weiner et al.), against an inventory of every fact said that relates to the patient's conditions, symptoms or signs (Weiner et al. 2020). Total per run, across 12 cases. The flat count is in Hallucination flags. | 107.6 | 89.0–121.0 | 3 of 6 |
| Errors reaching the record, per 100 notes (estimate)Note errors (Weiner et al. standard) per 100 notes, times 70%: in Biro et al. (2025), GPs reviewing AI drafts left each planted error uncorrected 65–75% of the time. An estimate of what a typical review would let through, across 12 cases. | 628 | 519–706 | 3 of 6 |
| Fewer errors than doctor-written notesCompared with the 5.0 errors per note Weiner et al. (2020) found in physicians' notes checked against recordings of the visit. Our test counts more strictly than theirs, so this favours the doctors. 100% means no errors; below 0% means more errors than the doctors. Across 11 cases. | -79% | -102%–-48% | 3 of 6 |
| Hallucination flags per note (Weiner et al. standard)Every clinical element the note documents that the consultation does not support, including wrong values, sides, doses and timings: Weiner et al.’s (2020) errors of commission. Paraphrase and clinical terms for what was said do not count. Average per note, across 12 cases. | 1.00 | 0.92–1.17 | 2 of 6 |
| Edit actions per noteMissing items, partial items and hallucination flags a clinician would need to fix, per note. | 8.97 | 7.42–10.08 | 3 of 6 |
| Clinical coverageShare of each consultation's fact inventory captured by the note (partial counts as half). | 81% | 77%–86% | 4 of 6 |
| Significant transcript errors per runMis-heard words that change clinical meaning, rated blind, across 12 cases. | 21.2 | 19.0–24.0 | 3 of 6 |
| Word error ratePooled word error rate per run against the human reference. Every word counts equally, including fillers. | 13.6% | 13.4%–13.9% | 3 of 6 |
| Stop to prompted noteMedian time from pressing stop to the note written to our instruction, per run. | 27s | 26s–27s | 4 of 6 |
| Data per consultAverage network data per consultation, per run. | 44.6 MB | 42.6 MB–47.5 MB | 6 of 6 |
| Avg RAM (stress test)Physical memory footprint (macOS `footprint`, matching Activity Monitor's Memory column) sampled every minute during one continuous 50-minute stress-test consultation — separate from the 12-case suite above. Every scribe recorded this visit at the same time on one machine, so absolute values reflect shared contention between them, not isolated real-world usage. Compare scribes to each other here, not to a device's spec sheet. | 1890 MB | 386 MB–2049 MB | 6 of 6 |
| Avg CPU (stress test)CPU used per minute during the same continuous 50-minute stress-test consultation, measured as cumulative CPU-seconds consumed divided by wall-clock time (can exceed 100% when a scribe uses more than one core). Every scribe ran concurrently on one machine. | 49.0% | 44.9%–53.0% | 4 of 6 |
Definitions and weights are in the methodology. Compare Heidi head to head ↗
26 January 2026
A clinician with 30 years in practice and a long-standing dislike of paperwork says Heidi clicked faster than the other tools they tried. They blame missed recordings on themselves and like the literature review added in the latest desktop update.
Summarised from a review on App Store. Not verified by ScribeStandard. Read the original
2 June 2026
After about a year in a busy outpatient clinic, this reviewer rates the product and support highly but says notes still need fixing. They report slow dictation that mishandles acronyms and dates, and recent bugs including a hot phone, fast battery drain and broken snippets.
Summarised from a review on App Store. Not verified by ScribeStandard. Read the original
25 June 2025
A free-plan user who once relied on the app feels updates have lowered note quality, sometimes below what they could write themselves. They also report notes failing to complete when wearing headphones, though the app has saved them a lot of time in the past.
Summarised from a review on App Store. Not verified by ScribeStandard. Read the original
9 December 2024
A freelance workers' compensation nurse case manager says Heidi saves significant time on complex cases, fits their workflow and helps with reports and communication between the parties involved.
Summarised from a review on App Store. Not verified by ScribeStandard. Read the original