Intensity Interferometry offers an interesting path forward for astronomical interferometry at very long baselines and at short wavelengths, all leading to extremely high angular resolution. Single photon counting detectors (and associated correlator technology) has evolved to allow a sensitivity gain of two orders of magnitude since the seminal experiments of Hanbury-Brown and Twiss in the 1960s, but sensitivity remains the challenge to make this technique widely useful: extremely high angular resolution requires high surface brightness by definition. The simplest way to increase the sensitivity of an intensity interferometer is to obtain multiple simultaneous correlation measurements at different wavelengths since these are uncorrelated and improve the SNR as √ N , with N spectral channels. The first issue to address is the break-even point since the throughput of spectrographs can be low; for example a throughput of 50% requires at least 4 spectral channel to break even. The next issue we wish to address is one of reliability and ease of use as we intend to deploy these spectrographs at multiple locations. In this paper, we propose and compare three concepts, each with its advantages and drawbacks: a classical multimode fiber-fed spectrograph, a photonics lantern fed focal plane based concept and an integrated optics solution using photonics lanterns and an Arrayed Waveguide Grating Spectrograph. We hope to build a simple demonstrator of whichever concept we end up choosing for testing in the context of the I2C project.