There are a lot of voices and I would need to test the voice and how what I want said sound before I export it. Is my character allotment used every time I click generate or only after I have my speech is configured and sounds the way I want it and it is exported? This could 1M would go fast if I have testing out the same sentence on 50+ voices.