I use a mixture of Metoffice, Yr.no and Weather Underground.
Overall, I have found this to be the most accurate combination for working out what is going on. I've tried plenty of others over the years and haven't found one single source that can accurately give you all the information you need.
I actually find that looking at the individual forecasts for lots of locations and building up a mental map is usually more useful than looking at rain radar maps. My standard procedure for a weekend in the peak is to start to look on Thursday and look at the forecast for a location on the Eastern Grit, Western Grit, Southern Grit, Eastern Lime, Central Lime and Churnet. I'll then narrow down locations if conditions are looking very marginal. I will occasionally look at radars after, but I've usually got all the information I need by then to know what is going on and which crags will have the best chance of good conditions.
Using this method, I have consistently been able to get out on weekends in good conditions on days when everyone else is complaining about the crap weather and resorting to training indoors. Even on this terrible winter, I climbed every single weekend except one where we decided not to bother with the drive. Two of those were days in slightly marginal conditions at Roche and Cademan but the rest were all good.
Useful for an overall idea of the weather. Not as accurate in the past couple of years as it used to be. The app is decent and it takes 30 seconds to get a good overview of all of my saved locations. Overall, it is slightly pessimistic in terms of rainfall.
Useful mainly for millimetres of rain. Will not give you info on things like fog/clag. Overall, it is slightly optimistic in terms of total rainfall expected. The app is decent if quite simple. The forecast doesn't have quite as much precision in terms of location as the others.
Its long term forecast uses a completely separate model to its hourly forecast. The long term forecast isn't at all accurate and isn't worth bothering with. Quite often, its long term forecast will be predicting entirely opposite conditions to those predicted by the Metoffice and Weather Underground (and, nearer the time, the yr.no hourly forecast). When the day comes around, it is nearly always the long term yr.no forecast that was wrong.
A caving focused forecast that gives you loads of detailed information. Great for working out which lime crags are likely to be dry on condensey days. The most pessimistic in terms of rainfall. The app is cumbersome, the UI sucks, but the info is useful.
Has the least inaccurate long term forecasts.
Adding in a quick browse of the UKC conditions page for the last 24 or 72 hours gives you a good idea of where has been dry too. If only they would let you save searches for the future.
I have in the past tried AccuWeather and BBC but their forecasts were rubbish, I might as well have just rolled a die and decided based on that.