Research · Guides · Methodology
Voice-list Blog
Research articles, methodology deep-dives and practical guides on speech-to-text software. Every claim is grounded in benchmark data from our test lab.
News
Wispr Flow accuracy dropped in June 2026 — the team explains why
Wispr Flow users reported a clear accuracy regression in June 2026. The team confirmed it and explained what changed. Here's what we know.
Research
Your dictation app's accuracy depends on which model you're using, not which app
We tested every model in every major dictation app. The same app can swing from 1.6% to 22.8% WER depending on which model is active. Here's what we found.
Guide
Which LLM should you use in Superwhisper?
Superwhisper lets you pair any voice model with any LLM. We tested the common combos — here's what actually works, and what the community learned the hard way.
Methodology
How we test speech-to-text apps
We run identical audio through every app and score transcripts on WER, CER and PER. Here's exactly how our benchmark works and why these metrics matter.
Comparison
Superwhisper vs Wispr Flow: which one is most accurate?
Benchmark-driven answer to the most common question about dictation apps. We ran both through identical audio tests — here's what the numbers actually say.
No posts in this category yet.