Radiofrequency ablation is a minimally invasive technique widely used for treating liver tumors, yet planning optimal probe trajectories for multiple tumors remains a significant challenge due to the complexity of avoiding critical structures and ensuring adequate tumor coverage. However, it is crucial to treat as many tumors as possible in one intervention to decrease patients’ hospitalization time. Furthermore, utilizing a single probe for ablating multiple tumors on the same trajectory reduces the number of probes used, and consequently, the risk of complications, such as crossing critical structures or trajectory collisions. In this scenario, probes are advanced to the most distant tumor, and subsequent to conducting ablation, the probes are retracted to the proximal tumor for an additional ablation. We propose a novel trajectory planning algorithm for ablation procedures, introducing an innovative multi-tumor planning strategy and force field-based navigation. Our genetic optimization algorithm is guided by a field derived from abdominal structures to enable efficient and safe navigation through complex anatomy. A retrospective analysis, performed on 18 patients from our in-house dataset, with 1 to 4 tumors each, shows its usability in clinical scenarios. Our algorithm produces safe, non-colliding, and clinically compliant solutions for all cases in 5.7 min on average and achieves a mean coverage of \(93.5\%\) of tumors with 5 mm safety margin. Comparison on single-tumor cases with existing automated methods demonstrates the competitiveness of our algorithm. Furthermore, the method’s ability to handle complex multi-tumor scenarios is a significant step toward clinical implementation.