Due to its outstanding merits, such as quick response, multiobjective optimization, and simple principle, model predictive control (MPC) has been widely used in power converters and motor-drive systems. However, MPC highly relies on the precise circuit parameters and control models, and cannot be used in unknown circuit relationships. To solve this issue, this article presents a model-free predictive control (MFPC) with multiobjective optimization (MOO) for two parallel three-level T-type rectifiers (3LT2Rs). First, the main control objectives of 3LT2Rs are analyzed, and the overall control scheme of the double closed-loop control is established. Second, based on the mathematical model of the parallel system, an MOO-MFPC for neutral-point voltage balance, current tracking, and zero-sequence circulating current elimination is proposed, which does not require any prior knowledge of the circuit parameters and circuit models, and it can achieve MOO control without weighting factors and its priority is not fixed. To solve the current difference updating stagnation problem in MOO-MFPC, a synchronous updating method is designed, which is faster than that of a single rectifier. Finally, the proposed method is tested on a hardware prototype of a 10-kW and a 5-kW parallel rectifier. Numerous experimental results demonstrate the merits of this method over the existing methods under several typical scenarios.