Kitaev's honeycomb lattice spin model is a remarkable exactly solvable model, which has a particular type of spin liquid (Kitaev spin liquid) as the ground state. Although its possible realization in iridates and α-RuCl3 has been vigorously discussed recently, these materials have substantial non-Kitaev direct exchange interactions and do not have a spin liquid ground state. We propose metal-organic frameworks (MOFs) with Ru 3+ (or Os 3+ ) forming the honeycomb lattice as promising candidates for a more ideal realization of Kitaev-type spin models where the direct exchange interaction is strongly suppressed. The great flexibility of MOFs allows generalization to other three-dimensional lattices, for potential realization of a variety of spin liquids such as a Weyl spin liquid.PhySH: Frustrated magnetism, Spin liquid, Quantum spin liquid, Organic compoundsIntroduction. -Quantum spin liquids, purported exotic states of quantum magnets where long-range magnetic orders are destroyed by quantum fluctuations, have been a central subject in quantum magnetism [1]. As an important theoretical breakthrough, Kitaev constructed a spin-1/2 model on the honeycomb lattice [2] with Ising interactions between spin components depending on bond orientations. Its exact solution demonstrates many intriguing properties such as fractionalized anyonic excitations. This model was later generalized to other lattices, including three-dimensional ones, still retaining the exact solvability [3]. In this paper, we call this type of model including various generalizations as Kitaev model, and its ground states as Kitaev spin liquids. Jackeli and Khaliullin [4] discovered that the "Kitaev interaction", namely bond-dependent Ising couplings, can be realized in a (111) honeycomb layer of iridates, i.e. the A 2 IrO 3 (A = Na, Li) structure, by the superexchange interaction through the oxygen ions due to the strong spin-orbit coupling of Ir 4+ in the Mott insulator limit (see also Ref.[5] for the itinerant limit).However, unfortunately, it turned out that iridates and related inorganic compounds, such as α-RuCl 3 [6], exhibit a conventional magnetic order at low enough temperatures and do not have a true spin liquid ground state. This is due to the non-Kitaev interactions, such as antiferromagnetic Heisenberg interaction, mainly coming from the direct exchange interaction between the metal ions [7]. While their finite-temperature properties still reflect the proximity to the Kitaev model [8] and thus are of great interest, the current situation calls for a more ideal realization of the Kitaev model in real materials, so that they exhibit spin liquid ground states.In this Letter, we propose such a possible realization of the Kitaev model in metal-organic frameworks (MOFs), crystalline materials consisting of metal ions and bridging organic ligands. Although MOF is a central subject in modern complex chemistry, MOFs have not attracted much attention in the context of magnetism. This is