Since i had my "Introduction to advanced macroeconomics" and discovered DSGE i've been wondering if IS-LM is used in research, i know it was used in the 80's or so, but now?
Yes, it is still used a recent example of IS-LM model used in literature is (Bonyah et al. 2019).
Also, it is important to understand those two are not substitutes for each other. IS-LM is a general equilibrium (GE) model. The way how you get dynamic stochastic general equilibrium (DSGE) model is that you take a general equilibrium model and add dynamism and uncertainty there.
As a student you were probably exposed more to the DSGE model based on real business cycle (RBC) model (which is based on Neoclassical GE models) as they are simpler, but New Keynesian DSGE models which are in essence DSGE version of IS-LM exists as well (see discussion of New Keynesian DSGE in Frische & Größl, 2010) and are relatively widely used (in fact I would say that after 2008 they became increasingly popular in the literature).
You will see DSGE models dominating over GE models in the literature, but this is because GE models are mostly used to derive some theoretical insights into how economy works, whereas DSGE can be used for actual practical simulations or even forecasting. You can't make any serious forecasts with theoretical GE model - no matter whether it is IS-LM GE model or some other one (only predictions about variables and relationship but that should not be confused with forecasting). Consequently, you will see DSGE models dominating because it makes sense to make separate paper about DSGE model calibrated for let's say economy of the Netherlands as in Lafourcade and de Wind (2012) (coincidentally they actually apply IS-LM based DSGE), whereas applying any GE whether IS-LM or some Neoclassical one to a particular country would not be very useful (at least not much beyond what you can learn from the general version already). So for every one theoretical GE model, no matter what variety, you can easily find more than dozen DSGE models.