To reduce the size, cost, and power consumption of current optical communication systems and to meet the requirements of future networks, it is crucial to develop modulators with high speed, low driving power, low loss, low cost, and small footprint. [1,2] Currently, the majority of available commercial modulators are based on silicon (Si), [1,[3][4][5] indium phosphide (InP), [6][7][8] and bulk lithium niobate (LN). [9] Despite impressive performance of modulators on these three platforms, none are capable of complying with the required criteria for the next generation of communication systems.In recent years, modulators based on thin-film lithium niobate (TFLN) have emerged as a promising approach for ultrahigh bandwidth modulators with low driving voltages and small footprint. [10][11][12][13][14][15][16][17][18][19] The optical modes in TFLN waveguides are more compact than conventional LN counterparts, allowing the radio frequency (RF) electrode spacing to be reduced without causing detrimental optical absorption loss. Consequently, the overlap between the optical and electric fields increases, resulting in a lower driving voltage. Furthermore, using a substrate with a lower dielectric constant than bulk LN makes velocity mismatch between optical and electric fields easier, and thus improving bandwidth. [20] Another factor that can enhance the optical bandwidth is increasing the thickness of the SiO 2 insulating layer under the thin films in order to optimize the velocity mismatch. [12] Despite the above inherent advantages of the TFLN platform, a delicate design is required to reach very high-performance modulators. For example, increasing the spacing between the electrodes can minimize the RF loss, a key parameter in determining the bandwidth, but at the expense of increasing the half-wave voltage, V π . [21] Various TFLN modulators on silicon substrate have been reported to expand the 3 dB bandwidth of a single modulator up to 100 GHz and higher. [11,12,22,23] One method is to use a large gap between the electrodes, which leads to a bandwidth of more than 100 GHz, but with a V π > 13 V, [11] due to the weak overlap between the electric and optical fields inside the gap. Another approach is to utilize asymmetric arms in a Mach-Zehnder modulator (MZM) and to operate at a null point of the transfer function. This approach, however, limits the operating wavelength and is sensitive to fabrication errors. [12] The present work proposes a novel design to relieve the tradeoff between bandwidth and V π in order to achieve modulators with ultrahigh bandwidth and reasonably low voltage-length