SCRIBE RESULTS / 22 SEPTEMBER 2026
Freed
Averages across 5 graded runs of 11 consultations, the same audio and instruction for every scribe. Notes graded on default template.
3m 7sTime to a signable note · ranked 7 of 7
20%Notes ready to sign as written · ranked 2 of 7
7.6Significant note errors per run · ranked 4 of 7
7.0Hallucination flags per run · ranked 5 of 7
24.2%Word error rate · ranked 7 of 7
Summary
Freed ranked 2 of 7 for notes ready to sign as written at 20% and 2 of 7 for clinical coverage at 92%. Its editing effort was 15%, ranked 3 of 7, with 1.98 edit actions per note.
Freed ranked 7 of 7 for word error rate at 24.2%, which ranged from 22.7% to 27.1% between runs. Hallucination flags averaged 7.0 per run, ranked 5 of 7, and ranged from 4.0 to 10.0.
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. | 3m 7s | 2m 59s–3m 15s | 7 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. | 20% | 18%–27% | 2 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. | 15% | 12%–18% | 3 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). | 2m 33s | 2m 28s–2m 41s | 7 of 7 |
| Worst note, any runThe single slowest note to review and fix across every case and run. | 3m 33s | — | 6 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. | 1m 52s | 1m 45s–1m 57s | 7 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. | 7.6 | 6.0–11.0 | 4 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. | 7.0 | 4.0–10.0 | 5 of 7 |
| Edit actions per noteMissing items, partial items and hallucination flags a clinician would need to fix, per note. | 1.98 | 1.73–2.18 | 3 of 7 |
| Clinical coverageShare of each case checklist captured by the note (partial counts as half). | 92% | 91%–93% | 2 of 7 |
| Significant transcript errors per runMis-heard words that change clinical meaning, rated blind, across 11 cases. | 29.6 | 27.0–31.0 | 5 of 7 |
| Word error ratePooled word error rate per run against the human reference. Every word counts equally, including fillers. | 24.2% | 22.7%–27.1% | 7 of 7 |
| Stop to prompted noteMedian time from pressing stop to the note written to our instruction, per run. | 46s | 44s–47s | 6 of 6 |
| Data per consultAverage network data per consultation, per run. | 16.5 MB | 16.0 MB–16.8 MB | 2 of 7 |
Definitions and weights are in the methodology. Compare Freed head to head ↗
Consultations
CONSULTATION