SCRIBE RESULTS / 22 SEPTEMBER 2026
Hanah
Averages across 5 graded runs of 11 consultations, the same audio and instruction for every scribe.
1m 28sTime to a signable note · ranked 1 of 7
45%Notes ready to sign as written · ranked 1 of 7
2.4Significant note errors per run · ranked 1 of 7
1.6Hallucination flags per run · ranked 1 of 7
12.7%Word error rate · ranked 1 of 7
Summary
Hanah ranked 1 of 7 for time to a signable note at 1m 28s. 45% of its notes were ready to sign as written, ranked 1 of 7, and its clinical coverage was 97%, also ranked 1 of 7.
Hanah averaged 1.6 hallucination flags per run, ranked joint 1 of 7, with a range of 1.0 to 2.0 between runs. Significant note errors averaged 2.4 per run and ranged from 1.0 to 4.0 between runs.
Every measure
| Measure | Average | Run range | Rank |
|---|---|---|---|
| Time to a signable noteWait from pressing stop to the graded note, plus review time: how long until the note can be signed if it is reviewed straight away. | 1m 28s | 1m 24s–1m 33s | 1 of 7 |
| Notes ready to sign as writtenShare of notes that captured every checklist fact in full with no hallucination flags, so nothing needed editing. | 45% | 36%–55% | 1 of 7 |
| 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. | 6% | 5%–7% | 1 of 7 |
| 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). | 1m 11s | 1m 9s–1m 12s | 1 of 7 |
| Worst note, any runThe single slowest note to review and fix across every case and run. | 1m 57s | — | 1 of 7 |
| Time saved per noteEstimated time to write the note yourself (5 s to recall each checklist fact, typing it at 40 wpm) minus review time. | 3m 15s | 3m 14s–3m 17s | 1 of 7 |
| Significant note errors per runMissed facts, partial facts and hallucinations that a blind AI rater judged could change diagnosis, treatment, follow-up, safety or the legal record, across 11 cases. Significance is a judgement call; the flat count is in Hallucination flags. | 2.4 | 1.0–4.0 | 1 of 7 |
| Hallucination flags per runEvery claim in the note with no support in the consultation, counted flat: each one counts, however minor. No severity judgement. Across 11 cases. | 1.6 | 1.0–2.0 | 1 of 7 |
| Edit actions per noteMissing items, partial items and hallucination flags a clinician would need to fix, per note. | 0.73 | 0.55–0.91 | 1 of 7 |
| Clinical coverageShare of each case checklist captured by the note (partial counts as half). | 97% | 96%–97% | 1 of 7 |
| Significant transcript errors per runMis-heard words that change clinical meaning, rated blind, across 11 cases. | 7.8 | 7.0–9.0 | 1 of 7 |
| Word error ratePooled word error rate per run against the human reference. Every word counts equally, including fillers. | 12.7% | 12.4%–13.3% | 1 of 7 |
| Stop to prompted noteMedian time from pressing stop to the note written to our instruction, per run. | 17s | 14s–22s | 1 of 6 |
| Data per consultAverage network data per consultation, per run. | 8.3 MB | 8.3 MB–8.5 MB | 1 of 7 |
Definitions and weights are in the methodology. Compare Hanah head to head ↗
Consultations
CONSULTATION