Search for a command to run...
Electro-optic (EO) transduction between radio-frequency (RF) and optical photons is a central building block for quantum networks, connecting remote superconducting and other solid-state qubits via itinerant photons. The strength of EO transduction is characterized by the single-photon coupling rate $g_0$, which quantifies how strongly a single RF and an optical photon interact with each other. Theory predicts that $g_0$ scales with the square root of the RF photon frequency and hence EO transduction is more efficient at higher frequencies, motivating the interest in extending electro-optic technologies from conventional microwave frequencies towards the millimeterwaves (30-300 GHz). This scaling aligns with recent trends to increase the qubit frequency into the millimeterwave regime to reduce thermal noise and relax cryogenic demands; however, the predicted $g_0$ frequency scaling has not yet been experimentally verified. Here, we experimentally demonstrate this scaling law using a triply resonant transducer on thin-film lithium niobate (TFLN) operating in the 90-370 GHz range. Millimeterwave signals are coupled to the chip wirelessly, via an on-chip dipole antenna that feeds a coplanar transmission line forming a Fabry-P\'erot-type cavity with loaded quality factors $Q\sim8-25$, enabling triply resonant transduction. This architecture supports the generation of electro-optic combs with line spacing in the millimeterwave and sub-terahertz bands. From the comb slope, we extract the single-photon coupling rate of $g_0/2\pi = $ 10.8$\pm$1.9 kHz at 370 GHz, which increases with frequency and is approximately five times larger than previously reported integrated electro-optic transducers. Our results establish wireless millimeterwave and terahertz cavities on TFLN as a practical route to high-$g_{0}$ EO transducers and highlight the appeal of developing qubits and quantum interfaces directly in the 100-400~GHz range, where both reduced thermal occupancy and intrinsically stronger EO coupling can be leveraged for scalable quantum networks.