Understanding the aqueous speciation of molecular metal-oxo-clusters plays a key role in different fields such as catalysis, electrochemistry, nuclear waste recycling, and biochemistry. To describe the speciation accurately, it is essential to elucidate the underlying self-assembly processes. Herein, we apply a computational method to predict the speciation and formation mechanisms of polyoxovanadates, -niobates, and -tantalates. While polyoxovanadates have been widely studied, polyoxoniobates and -tantalates lack the same level of understanding. First, we propose a pentavanadate cluster ([V₅O₁₄]^3–) as a key intermediate for the formation of the decavanadate. Our computed phase speciation diagram is in particularly good agreement with the experiments. Second, we report the formation constants of the heptaniobate, [Nb₇O₂₂]^9–, decaniobate, [Nb₁₀O₂₈]^6–, and tetracosaniobate [H₉Nb₂₄O₇₂]^15–. Additionally, we compute the speciation and phase diagram of niobium, which so far was restricted to Lindqvist derivates. Finally, we predict the formation constant of the decatantalate ([Ta₁₀O₂₆]^6–) in water, even though it had only been synthesized in toluene. Furthermore, we also calculate the corresponding speciation and phase diagrams for polyoxotantalates. Overall, we show that our method can be successfully applied to different families of molecular metal oxides without any need for readjustments; therefore, it can be regarded as a trustworthy tool for exploring polyoxometalates’ chemistry.