AXL is a receptor tyrosine kinase that has been shown to be a major player during tumor metastasis. Recently, many studies have defined AXL as a therapeutic target for metastatic disease. However, despite several therapeutic approaches including small molecules and antibodies have been developed to inhibit AXL signaling, only moderate therapeutic efficacy in metastatic in vivo models has been reported. In this study, by using directed in vitro evolution approach, we engineered a “super-binder” AXL mutant protein with a superior binding affinity to the only known AXL receptor ligand GAS6, thus preventing ligand-mediated activation of AXL receptor signaling. Compared to wild-type AXL/GAS6 binding, the super-binder mutant has a enhanced improvement in the binding affinity to GAS6. When tested in vivo, this improvement in binding affinity was translated into a significantly improved antitumor responses and survival in many metastatic tumor models including the SKOV3 and OVCAR8 metastatic ovarian models as well as the 4T1 spontaneous metastatic breast cancer model. Subsequent analysis also showed reduction in pAXL and pAKT expressions in tumors treated with the super-binder, indicative of successful inhibition of the AXL signaling cascade. Furthermore, no tissue toxicity was observed in animals treated with the super-binder protein. Collectively, we have engineered a mutant AXL protein with enhanced GAS6 binding affinity and improved therapeutic outcome, which has implications for a number of metastatic diseases that currently lack effective treatments.