I'll chip in with Russia. I expect the below figures to be massively higher for Western nations (which achieved near-universal literacy in the late 19th C) but I'll leave those to an expert.
Tsarist Russia had a slightly ill-deserved reputation for backwardness. On the one hand, its population was unquestionably relatively uneducated: the 1897 census reported that only 21% of the population was illiterate, far below Western norms. Yet this was changing: there was decent investment in education during the late Tsardom and by 1914 it's been estimated that the literacy rate in Russia was around 40%.
That single figure masks a whole range of variables though. There were marked differences based on generations, geography, urban living, and gender. For example, using the more reliable 1926 data, European Russia was significantly more literate than Central Asia, the urban world was 20-30% more literate than the rural, and about twice as many men could read as women (in rural areas, less so in towns/cities). These trends would have broadly held true in 1914.
The actual literacy rates from the 1926 census are:
All of which is civilians. Given that the war generation was relatively well educated (and that soldiers are obviously male) the recorded figures are that 68% of soldiers from working class backgrounds were literate, with the corresponding figure for peasant soldiers being 39%. The vast majority of soldiers were peasants and Boris Mironov thus estimates that some 43% of the Imperial recruits were literate.
So about 40% of civilians and soldiers in Russia 1914 were literate, albeit with major discrepancies between demographics.
Sources: The 1914 figures I got from Davies et al's Economic Transformation of the Soviet Union, while the 1926 numbers came from Fitzpatrick's Education and Social Mobility in the Soviet Union (see also her Everyday Stalinism). Mironov's The Standard of Living and Revolutions in Russia is the source for the military/recruit data.