This paper analyses the various ways of carrying out near IR multi-object spectroscopic studies in space. We show that ground-based observations would have limited results except in the 1-1.5 micrometers wavelength where large telescope of the 8m class would be approximately equivalent to a 1m in space. Beyond 2m, even an instrument such as the adaptive-slit near IR (ANIS) would be much more efficient. Due to their position in space, the traditional masks used in ground- based telescopes cannon be used. New technologies must be developed. Here, we present a multi-object spectrograph called ANIS based on micro-mirror arrays and designed for NGST PathFinder3. It would be able to perform a near IR spectroscopic/photometric mini-survey of the sky over a few square degrees. Thanks to its large field of view, ANIS would be complementary to NGST. Its goal would be to probe the Universe in the 0 < z < 5 range and we can consider ANIS as a scientific precursor for the NGST.