Wispr Flow

I used Wispr Flow as my primary dictation tool for months and it earned that spot. Straight up: it beat everything else I had tried, by a lot. If you are on a platform it supports and you write by talking, it is the best thing in this category and I would still recommend it.

What it does well: the accuracy is the point. It works system-wide into any app. The auto-editing and tone matching are real, it strips the false starts before the text lands. And it's polished, you can tell it's a funded product with people on it.

Two honest problems, different sizes. The small one: it's a subscription, $12 to $15 a month. Worth it if you dictate daily, which I did.

The larger one is architectural, and it's why I eventually left. It's a closed system. Your audio leaves your machine and the transcription happens on infrastructure you don't own or control. They say they handle it responsibly and I have no evidence otherwise. But it's a dependency, in the exact place where I stopped being willing to have one.

On iPhone, the Action Button integration is the weak spot. I used it on my phone and it never felt like it belonged to the OS. It needs tighter integration, and here's the thing: that's not really their fault, it's iOS. Apple doesn't hand third parties the press-and-release events you'd need to make a hardware button behave like a real hold-to-talk key. So every dictation app on iOS is working around the same wall. Worth knowing before you buy based on the button.

And the practical problem for anyone on Linux: there is no Linux build. Not a beta, not a workaround. I moved my whole desktop to Linux this year and Wispr Flow could not come with me. That's not a knock on them, it's a platform decision, and they may change it. But if you're on Linux today, this isn't an option, and the page here should say so.

I still think it's a genuinely good product. I just needed it to run where I work.