Translating screen captures can be done without needing to use the ztranslate client directly. This can be useful for cases where you want to translate screen captures yourself, or if you want to use a translation integration available in an application. For this case, you can use a project like vgtranslate (http://gitlab.com/spherebeaker/vgtranslate), which will run a local server that will contact apis to translate for you, or you can use the ztranslate.net service endpoint.
An example request would be a POST request to:
https://ztranslate.net/service?target_lang=En&source_lang=Ja&mode=Fast&output=png,wav&api_key=< api key >
With body:
{ "image": "Bas7G... <base64 encoded image data> ...34==" }
If target_lang is not provided, it will assume english as the target language, if source_lang is not provided it will guess the input language, if mode is not provided it will default to "normal" mode and save the image to the user account's uncategorized images. output is a comma separated list of formats to output. Supported formats are most image formats, and wav for Text-To-Speech output. If both an image and sound output is given then it will output both in the request in the "image" and "sound" fields. If none is given, it will output an bmp image. If "image" or "sound" is given, then it will return a bmp and wav respectively. api_key is currently optional, but highly recommended.