Superwhisper
Local Whisper dictation for Mac and iPhone, with optional cloud models
Dictation app for Mac and iPhone that runs Whisper models locally by default, with optional cloud models and bring-your-own-key API access on the paid Pro tier. Free tier available for local use. Pro is offered as a subscription or a lifetime purchase.
Index summary: Superwhisper scores 6.0 ± 0.2 out of 10 on The Dictation Index, rank #9 of 12 and statistically tied with ranks 9 to 10, from 554 public posts on X dated Aug 21 to Sep 16, 2026.
Where Superwhisper wins and loses
Most favorable: Speed (39 positive, 6 negative). Least favorable: Reliability (2 positive, 15 negative). Topics with fewer than 15 labelled posts are marked as thin and left out of that comparison.
| Topic | Posts | Positive | Negative | Neutral | Balance |
|---|---|---|---|---|---|
| Usability | 108 | 85 | 22 | 1 | |
| Accuracy | 65 | 39 | 24 | 2 | |
| Platform support | 60 | 18 | 30 | 12 | |
| Price | 55 | 19 | 33 | 3 | |
| Speed | 47 | 39 | 6 | 2 | |
| Privacy | 39 | 30 | 5 | 4 | |
| Reliability | 17 | 2 | 15 | 0 | |
| Support thin | 14 | 4 | 8 | 2 |
A post can discuss several topics or none, so these counts do not add up to n = 554. Counts are published in tools.json.
What people say
Example posts from the sample, each linked to the original on X and edged with the label it was given. A handful of posts cannot stand in for 554, so read the score and the topic table first. Quoted here and want it removed?
Willow annoyed me by not supporting Hebrew, so I moved to Wispr Flow for half the price. Then Wispr killed the old Pro plan, so I moved to Superwhisper. AI dictation is a commodity so loyalty lasts exactly until the next annoyance.
Some examples of what i use @grok 4.6 for. Im replacing software i used to pay for and use every day. - Radio channels + podcasts with a Spotify looking interface the video) - Vivid for my Macbook m4 max - My own Superwhisper but Norwegian trained, works a lot better than the
S1-Voice by @superwhisper feels near instant. What a model for dictation!
previously, superwhisper charges a subscription fee even when you just needed to run a _local model_ on your _own hardware_ that’s just blatantly wrong, and was why i chose to use open source solutions you are not “making whisper models free”. it’s already free. you just
i really need @superwhisper to start including my tone of voice when i'm dictating to LLMs because i don't think claude is catching on with how disappointed i am.
If you use https://t.co/bFB7NC2Ndd for speech-to-text, don't forget to enable post-processing! You can even use Apple's native models to remove any blabber you had while speaking! Idk why people use Wisprflow/Superwhisper when this piece of fine software is open-source!
Registry facts on this page were checked against Superwhisper's own site on 2026-09-16. See something wrong, or represent Superwhisper? Request a correction or send a reply. Comparing options? See alternatives to Superwhisper.