Lithium metal anode (LMA) is considered the most promising candidate for energy-dense batteries and is widely employed for its extremely high gravimetric capacity (3860 mA h g⁻¹) and volumetric capacity (2060 mA h cm⁻³) and the lowest redox potential (−3.04 V vs. SHE). However, the commercialization of LMA has been prevented by the interior lithium morphology instability and high N/P ratio on the exterior. Lithium morphology instability leads to poor cycling performance and security risks, while the high N/P ratio results in the low true specific capacity of LMAs and low energy density of pouch cells. In order to solve these two issues, we developed a thickness-controllable lithium–zinc composite anode consisting of β-Li and the intermetallic LiZn compound by roll-to-roll pressing, which allowed the preparation of a thin Li–Zn foil (20 μm). The rod-like LiZn compound acted as a 3D mixed-ion and electron conductor (MIEC) skeleton that offers a large number of binding sites for Li adatoms, thus achieving morphological stability, less consumption of the liquid electrolyte and cyclable lithium. On this foundation, the Li–Zn composite achieved better electrochemical performance during the plating and stripping process with a lower overpotential. Furthermore, coin cells with a LiFePO₄ cathode and the Li–Zn anode exhibited 1.6 times the cycling life of those with LMAs. The high-capacity NCM811 pouch cell with an N/P ratio of 1 achieved 50 stable cycles with the Li–Zn anode, and the energy density was up to 357 Wh kg⁻¹ and 1223 Wh L⁻¹.