You should discuss this with your local gas vendor. If high purity gases are not required for your work of analyzing alcohol purity, for example, then pricing of nitrogen and hydrogen may vary considerably.
But, if you require high purity gases then hydrogen generators may be the required solution.
As far as use, it is far easier to get comparible if not improved results with hydrogen carrier gas. Nitrogen is more tricky to use, especially if you wish to keep the separations you have achieved in the past.
The real issue is cost, and that was the basis of your question. But location and availability of the gases are more important than a 'general' answer.
If you use FID as your sole detector then you have to supply hydrogen anyway, so why bother with nitrogen?
Your capillary use of hydrogen CAN be less with the column than with the detector itself, which generally requires a 35-45mL/min supply. A capillary splitter may require 60 to 120mL of flow, or perhaps much less if a megabore column with direct injection is used (10mL/min or less).
Based on my experience and the possible need in the future for high temperature or high resolution using 0.18 or 0.10mm ID columns, I would pick hydrogen.
best wishes,
Rod