In　this　paper,　a　new　technique　is　presented　for　optimizing　microwave　tunable　filters.　The　technique,　which　is　based　on　surrogate　models,　addresses　the　challenges　of　meeting　multiple　sets　of　tuning　specifications　simultaneously.　By　employing　a　collection　of　ideal　solutions　for　fixed　parameters　and　several　sets　of　ideal　solutions　for　adjustable　parameters,　the　proposed　technique　achieves　this　goal.　The　surrogate　model　is　made　up　of　several　sub-surrogate　models,　each　with　inputs　of　non-tunable　and　tunable　parameters.　The　complete　surrogate　model　has　inputs　of　non-tunable　parameters　and　multiple　sets　of　tunable　parameters.　By　optimizing　all　the　design　specifications　together,　the　technique　being　suggested　is　more　likely　to　discover　appropriate　solutions　that　meet　all　the　requirements.　Moreover,　the　method　is　more　proficient　in　steering　clear　of　local　optima　and　achieving　the　optimal　solution　with　greater　efficiency　compared　to　the　current　two-stage　optimization　approach　used　for　distinct　specifications.　The　validity　of　our　technique　is　demonstrated　using　a　tunable　four-pole　waveguide　filter　optimization.　©　2023　IEEE.