Although visual perception of humans is limited to a fraction of wavelengths spanning the electromagnetic spectrum, technological advances enable us to see in other spectral regions by providing suitable sources and detectors. Of particular interest for many applications is the ability to probe objects in the terahertz (THz) range, which bridges the microwave and infrared domains. THz radiation offers unique opportunities for imaging or sensing due to its high transmission by optically-opaque materials like paper, textiles, ceramics or plastics, while for gas sensing it enables identification of structurally-complex molecules. Unfortunately, access to this region is difficult due to limitations of conventional electronics and photonics, and often involves cryogenic operation. Even the most mature systems operating at room temperature, despite years of advances, still struggle to provide chip-scale miniaturization of the source and detector, and moving-parts-free acquisition of a broadband THz spectrum. Here, to fill this niche and address the critical demand for broadband, chip-based THz spectroscopy without any moving parts, we propose to leverage mid-infrared (3-5 um) semiconductor laser frequency combs based on interband cascade lasers (ICL). We postulate that nonlinear frequency conversion due to the recently discovered second order susceptibility of the ICL medium can be used to obtain microwatt to sub-milliwatt level of THz power at a battery-compatible bias. A complementary mid-infrared photomixer technology envisioned in this proposal will additionally enable coherent detection of broadband THz comb radiation at room temperature. Although the project is inherently risky due to uncertainties in the ultrafast dynamics of semiconductor structures, losses in the terahertz range, and fabrication complexity, it is timely and strongly demanded by the community. It will unlock new opportunities across many disciplines ranging from chemistry to 6G telecommunications.