The standard LCDM model gives a successful description of many astrophysical observations. However, in the past few years a tension has developed between local determinations of the Hubble constant and the value predicted from early universe probes. If confirmed, this so-called Hubble Tension, would require additional physical ingredients beyond LCDM, e.g. early dark energy, or new particles. After describing the tension, I will provide an update of a 25-year long effort to measure the expansion history of the universe and thus the Hubble constant using gravitational time delays, highlighting recent results based on lensed quasars from the TDCOSMO collaboration (the union of H0licow/STRIDES/SHARP collaborations), and from multiply imaged supernovae, including Refsdal. I will show new JWST HST Keck and VLT data that we are currently analysing with the goal of achieving 1-2% precision and accuracy on H0 and thus contribute to the resolution of the Hubble tension.