SovereignCopper "Raw, uninterpreted data" is perhaps a bit too harsh. There's are several models that analyze the data which can be chosen from via the APIs. Additionally, the precipitation forecasts were much better in the past, at least in my experience.
Sorry, I really meant "raw = not analyzed by a human specialist". But yes, there are apps that display weather info exactly as provided by the APIs with no further processing. Nowadays with AI, it's possible to get some basic interpretation of the results but still not as good as a human, at least for now.
As for the forecasts being better in the past, there's a reason for that. In the past, global, per region/country, forecasts were good enough for everybody but nowadays, what matters is very precise forecasts almost to km²/miles² places for specific uses like agriculture, fishing spots, sensitive places, etc. Region/country wide forecasts for the general public does not need to be that precise. The latest models used are so complex and require so much processing power that only the pinpoint accuracy of specific places are prioritized. At least, that's what happens in my country according to various documentaries.