Smallest Pulse Author Benchmark - mode not reported - Private English perturbations - Disfluency - WER: leaderboard
Smallest Pulse Author Benchmark - mode not reported - Private English perturbations - Disfluency - WER — Smallest AI Pulse own author model-card evaluation, content hash pinned. Native model=pulse; never Pulse Pro or Pulse2. Comparator observations with unbound product versions retained as evaluation runs, not public model releases. Native percentage values unchanged; error rates unbounded above100. mode not reported. Private author audio sliced by perturbation; this is not public WildASR. Metric: WER (%). Source: https://docs.smallest.ai/models/model-cards/speech-to-text/pulse
Metric: WER (%). Source: docs.smallest.ai. Saturation forecast: Rough model projection: around 2026. 5 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Assembly Universal 3 Pro | smallest_pulse_author_benchmarks run | 7.81 | |
| 2 | Smallest AI Pulse | 8.2 | ||
| 3 | Deepgram Nova 3 | 8.62 | ||
| 4 | AWS Transcribe | smallest_pulse_author_benchmarks run | 9.23 | |
| 5 | ElevenLabs Scribe | smallest_pulse_author_benchmarks run | 9.29 |
Interactive version: theaggregate.ai/benchmark?slug=smallest-pulse-author-benchmark-mode-not-reported-private-english-perturbations-disfluency-wer · How It Works · Data refreshed daily, snapshot 2026-11-16.