Is monte carlo supported by the analogue models at the moment? I see lots of statistical values interspersed in the models but it's not clear how I would set up a MC run. Looking at the ngspice example you need to pick a random value for each parameter in a loop. For real models (not text book models like the example) this could be a lot of manual work to find all the parameters and add them to the loop. @User I've looked at your example comparator which gives some insight but it only seems to vary only two parameters, is this sufficient? Also for mismatch? It may be this is just not really supported yet but I wanted to check