Agrivoltaic systems are expanding in semi-arid regions, raising questions about how panel driven shading alters soil greenhouse gas exchange. We used automated chambers to measure high frequency soil CO2, N2O, and CH4 fluxes in a newly established vineyard agrivoltaic system in Texas across three shading intensities (0%, 35%, 88%) that represent the light environment beneath panels as shaped by panel tilt and solar trajectory. Shading primarily cooled the soil, reducing mean soil temperature from 35.9 degrees C (0%) to 32.5 degrees C (35%) and 30.5 degrees C (88%), while mean soil water content shifted modestly (0.125, 0.106, and 0.130 m3 m-3, respectively). Over three months, cumulative soil CO2 emissions declined from 3.2 Mg C ha-1 (0%) to 2.6 (35%) and 2.2 Mg C ha-1 (88%). N2O emissions were dominated by rainfall driven pulses and were lowest under 35% shade (0.15 kg N ha-1) without increasing under 88% shade relative to full sun (similar to 0.17 kg N ha-1). All soils were net CH4 sinks, with greater uptake under shaded treatments (-0.78 to -0.77 kg C ha-1) than under full sun (-0.24 kg C ha-1). In CO2 equivalents, shading reduced net emissions by 0.62 and 1.02 Mg CO2 eq ha-1 under 35% and 88% shade, corresponding to 18.9% and 31.2% reductions. These results show that agrivoltaic shading can lower net soil GHG forcing in semi-arid viticulture by buffering soil temperature extremes, suppressing CO2 losses, and strengthening the methane sink without increasing cumulative N2O emissions.