The option -dump dumps the raw API response (almost always JSON) to the stdout for future consumption (by a 3rd party program):

$ trans -dump :fi 'hello world'
[[["Hei maailma","hello world",null,null,1]],null,"en",null,null,[["hello world",null,[["Hei maailma",1000,true,false],["hello maailma",0,true,false],["Hello World",0,true,false]],[[0,11]],"hello world",0,0]],1,null,[["en"],null,[1],["en"]],null,null,null,null,null,[["World"]]]

The option -download-audio downloads the Text-to-Speech audio into the current folder (#77). It by default uses the translated text; when used in combination with -speak or -no-trans it yields the audio of the original text. (Known issue: -no-trans doesn't work with Bing engine.)

$ trans -download-audio :fi 'hello world'

Alternatively, -download-audio-as may be used to specify a filename:

$ trans -download-audio-as=output.ts :fi 'hello world'