It's free, it looks great and it's got a lot of knobs. For this reason it was one of the first compressor plugs I ever tried out. A couple of years later and I'm still astonished by this thing. Don't think because it's free that it's in some way inferior to a compressor with a huge price tag. This is up there with the most advanced software compressors ever created.
Variable attack envelopes, mid/side effects, high pass on the sidechain, optional warm saturation, knee shape, upsampling and parallel. There really isn't an option left out. It's also the only compressor I've ever seen that has a mid-scoop eq. Not sure why you'd really need one built into the compressor but I've come to like it being there.
One thing it won't do however is a huge squish. Even when I max out the ratio and threshold it remains fairly civilized and only reduces around 20db. But I guess no sensible person would ever really want to do more than that.
Two hints for noobs:
a)The alpha/sigma switch changes the attack shape. Alpha is more standard and sigma lets most of the original signal pass through the attack. No matter how many times I use Molot I still seem to forget which is which.
b) The limiter has two possible routings: "Pre" - compressor -> limiter -> EQ -> Makeup -> Dry Mix, or "Post" - compressor -> EQ -> Makeup -> Dry Mix -> limiter. Don't get confused like I did and wonder why the heck anyone would want to use a limiter before the compressor. To make things worse there is unfortunately a bug on the latest version (3.1) that means the limiter is always set to the "pre" routing. Hopefully we see a fix for this soon.
The only real problem you might have with Molot is that it is in fact too big and versatile for your needs. Sometimes it's simpler to reach for a compressor with just a few knobs.
An inspiring and artful plugin... definitely worth clicking on the donate button...
PS: The VU meter isn't without a few idiosyncrasies! It does a good job of GR but it doesn't read the input very precisely. On the 2.168 version it does a passable job but it certainly isn't reading the peaks and occasionally it reads well above. On the 3.1 version it is considerably worse and often reads way way above. This could be a bug or it could be an alternative approach to modelling. Either way it's not very useful. Thankfully the input meter is something you can easily do without. The GR meter seems to be much more accurate (on both versions). It gets the db acceptably close, bounces in the right places and is very easy to read.