alt Hacker NewsI find the whole category of "added value" STT apps funny at this point.
Built my own in a day that blows them out of the water. You can quite literally pick any sufficiently good local model or API provider and combine it with Cerebras for cheap and very fast AI post processing and formatting.
If you want the best accuracy, then this is not true. Combining models will beat any single model. There's also a lot of work that goes into optimizing dictionary words and screen context, and knowing which ones to, to add to model context.