I had purchased a lifetime license on another tool I had used for it's voice cloning and it went out of business and AppSumo gave me the money I spent back in credits under their guarantee. So I went all in on this voice cloning online app and went with the highest tier.

VoiSpark is definitely better at voice cloning and text-to-speech than the other app was. It is really intuitive to use and easy to store and recall several custom voice clone profiles. I used it right away for a simple non-fiction video and that worked great. Then I tried it for a project where I cloned my voice several times doing different voice acting characters for reading fiction stories that we dub over simple video animations. I used VoiSpark to help me stay consistent between characters with bits of dialog. Even though I would love to not have character limits (for longer projects with consistent voicing), for me a 5k character limit at a time hasn't been a big issue and is actually just right to make sure I like the lines I want to give to my editor without rendering out too much at a time and wasting credits. Since I had the top tier, I wasn't worried about the one-time 100k credits used to clone each voice since I could store those profiles with descriptions and use them again later for other projects later - and it would then be cheap to recall the voice profile after that first use. With the max tier I didn't feel like I was in danger of running out of credits anyway.

I like the "happy' emotive setting on MiniMax for a lot of different kinds of general readings and toying with other settings in addition to "emotion" to get shades of characters that still sound like I voiced them but varying in pacing or nuanced differently. If I don't like the way it was read I found if I just ran the lines through again and played with the punctuation or tweaked anything, the result was a noticeably different read that I often liked better. One suggestion: The ability to save different setting tweaks separately for a single cloned voice would be great.

I'll admit I revealed a bug in the first hour of using It, but things have been smooth for the most part. It just had issues with reading chapter headers or odd punctuation that humans would know how to read but an AI model is still unsure of. For example, I had something like "Book 3.9.1: Section Title" that didn't read correctly at all (it actually said some gibberish after the word "book" that didn't make sense) and I had to go in and change it to something like "Book one point five point two. Section title" and it worked much better. I didn't notice how the "text normalization" setting helped with stuff like this but it is a setting that the app says is supposed to help - so I kept it on.

I was able to do samples of minor, but difficult, voice acting characters that were hard on my voice for me to keep up, or infrequently used in the story and difficult to recall and conjure up consistently when there were large gaps between their appearances within the story, or just difficult to voice in general. I used VoiSpark to help quickly recall and perform these difficult character voices. Having it there in a saved character list to automatically refer back and perform lines for me made it much easier to save my voice and remain consistent with that character for an entire book or episode.

I'm giving it five stars because my creative team is using it every day. I am confident enough with the tool to let the editors render lines out for me to save my voice acting time and it is doing a great job for the way we use it. I really want to see this company stick around for the long haul. It's missing some conversational/podcast settings that was in one of the other voice cloning apps I've used (and I do miss that). And I'd love for more settings to tweak cloned voices and more tools for using it to help voice custom characters believably and help tell fictional stories. All in all, I think it's a great deal and I love having a lifetime deal in place for a good voice cloning tool like this. Please don't go anywhere, guys!