A decanuclear silver chalcogenide cluster, [Ag10(Se){Se2P(OiPr)2}8] (2) was isolated from a hydride-encapsulated silver diisopropyl diselenophosphates, [Ag7(H){Se2P(OiPr)2}6], under thermal condition. The time-dependent NMR spectroscopy showed that 2 was generated at the first three hours and the hydrido silver cluster was completely consumed after thirty-six hours. This method illustrated as cluster-to-cluster transformations can be applied to prepare selenide-centered decanuclear bimetallic clusters, [CuxAg10-x(Se){Se2P(OiPr)2}8] (x = 0–7, 3), via heating [CuxAg7−x(H){Se2P(OiPr)2}6] (x = 1–6) at 60 °C. Compositions of 3 were accurately confirmed by the ESI mass spectrometry. While the crystal 2 revealed two un-identical [Ag10(Se){Se2P(OiPr)2}8] structures in the asymmetric unit, a co-crystal of [Cu3Ag7(Se){Se2P(OiPr)2}8]0.6[Cu4Ag6(Se){Se2P(OiPr)2}8]0.4 ([3a]0.6[3b]0.4) was eventually characterized by single-crystal X-ray diffraction. Even though compositions of 2, [3a]0.6[3b]0.4 and the previous published [Ag10(Se){Se2P(OEt)2}8] (1) are quite similar (10 metals, 1 Se2−, 8 ligands), their metal core arrangements are completely different. These results show that different synthetic methods by using different starting reagents can affect the structure of the resulting products, leading to polymorphism.