Verified purchaser
Wow!
I've used Blip AI and other various dictation tools on AppSumo, and they all have their benefits and downfalls. However, Amical is by far the most accurate and fastest among all of them. I had some issues yesterday when I initially started using it, but afterwards, it's all smooth sailing.
What I like about Amical is that it is accurate, it is fast, and for the most part, it has the least amount of issues. When I used BlipAI, I had to update to the software whenever I had an issue or it just doesn't work whatsoever. But I have to admit that BlipAI has been the longest diction tool I've used. Granted, Amical does not have the AI feature like BlipAI. However, for the most part, I don't really need that feature, nor do I use it that often. As well, I have another software that edits my text when needed. P.S. I've used Amical to "write" this review, minus some edits on my part that I had edited afterwards.
I had an issue with one diction while writing this review. Error Message: transcription failed. However, I really appreciate that this tool has the audio version of what I just said and that I can download it, so that if there is an issue, I don't have to reinvent what I just said; I could just extract the key points.
Initially, when I started testing this out yesterday, when I used the manual key shortcut, it stopped me from continuing and processed the two or three seconds of text that I just dictated. However, when I use the continuous shortcut, it also did that on Windows. However, the bottom icon, when I clicked on that, the issue was gone. But, now I just have to click the continuous dictation and it works fine.
I'll have to do more testing, but I think that was when I was using the local models.
Haritabh_Amical
May 21, 2026Thanks so much for the detailed review. We have invested a lot into the core dictation experience, accuracy & latency, and hearing it praised makes our day :)
Sorry about the issues you ran into initially. Local models can be a bit more variable because performance depends heavily on the machine, drivers, available memory, and other system-level factors. We know the UX around this can be better, and we are working on improving how Amical handles and communicates local-model issues.
If you suspect the failed transcription happened while using the cloud model, please report that specific dictation from the History page. You should see a flag icon when hovering over the entry, but only for cloud transcriptions. That gives us the context we need to investigate properly.
Also, one useful note: the audio is not just saved locally for download. You can also retry a failed transcription from the History page. So if a local model fails or gives poor output, you can switch to the cloud model, go back to History, and retry the transcription from the same locally saved audio without having to dictate it again.
And if you have some time, we would love to pick your brain on two things: how the other software you use edits your text, and whether something like that should belong natively inside Amical. We would also love to better understand the shortcut issue you ran into so we can improve that flow.
You can reach out to us directly through our community (discord server) at https://amical.ai/community or by email at [email protected].
Really glad to have you onboard, and we’d love to keep getting feedback as you continue testing Amical.