In ongoing attempts to synthesize nanosized gold-platinum carbonyl phosphine clusters either directly or indirectly by initially obtaining intermediate-sized ones as potential precursors, the neutral Au–Pt CO/PR3-ligated cluster, Pt7(μ 2-CO)8 (PPh3)4(μ 4-AuPPh3)2 (1), was isolated and characterized by low-temperature CCD X-ray diffraction, IR, and mass spectrometric measurements. The heretofore unknown 9-atom metal architecture of the Au2Pt7 core may be envisioned as a Pt7 adduct of a Pt4 butterfly and an edge-opened Pt3 triangle that is additionally linked by two tetracapping AuPPh3 units. Each of the two butterfly-hinged (basal) Pt atoms is connected to two of the three edge-opened triangular Pt atoms thereby giving rise to an essentially coplanar Pt5 fragment consisting of three edge-fused bonding triangles; addition of two AuPPh3 units to the Pt7 adduct completes the Au2Pt7 framework via formation of the two symmetry-equivalent square-pyramidal (μ 4-Au)Pt4 moieties, each resulting from the tetracapping of a AuPPh3 unit to the central Pt3 triangle of the coplanar Pt5 fragment and to one of the two out-of-plane wingtip Pt atoms. Six PPh3 ligands are coordinated to the two Au atoms, the two wingtip Pt atoms of the Pt4butterfly-shaped fragment, and the two outer Pt atoms of the edge-opened Pt3 triangle. Four of the eight doubly bridging μ 2-COs link the four edges connecting the two out-of-plane wingtip Pt atoms with the two hinged (basal) Pt atoms, while the other four μ 2-COs span four of the five Pt–Pt bonding edges of the coplanar Pt5 fragment (i.e., the fifth Pt–Pt bonding edge is spanned by the two out-of-plane wingtip Pt atoms). The resulting molecular configuration (without P-attached phenyl substituents) possesses crystallographic C2 (2) and pseudo-C2v symmetry. An optimized geometry of the PH3- and PMe3-models of 1, obtained from gradient-corrected (scalar-relativistic) DFT calculations is in reasonably good agreement with the crystallographically determined geometry of 1.A structure-to-synthesis approach is presented as part of extensive but unsuccessful efforts to design a high-yield reproducible preparative route to 1 in order to enable physical/chemical studies as well as to utilize it as a preformed cluster for further PPh3/CO deligation reactions.