LaunchPod AI Reviews

I love the realistic voices and scriptwriting abilities. A+

When I first tried this about a month ago I wasn't very impressed, the app repeated some words and didn't sound all that great. Plus there was a pretty small word count limit.

I just tried again today as I heard the audiobook feature was released. This is significantly better than it was the first time I tried the app. I copy and pasted 1430 words into the audio book portion and selected the pre made voice that sounded the least robotic. I don't recall any repeated words, the inflections are mostly right but not quite there yet. Some parts still sound a little robotic. If I were to compare this to eleven labs I would say this is about 70% of the way there....at least for the best default voice I tried.

With this being a new product I don't expect perfection right now, and with the audiobook portion I'm pretty happy with the direction this is going(this is actually about the level I was hoping for at launch).

Lets put it this way, a good narrator will beat this right now, but a good narrator would also beat eleven labs.

I've turned off audio books before because I thought the narrator was really bad. If LaunchPod was used on those few books I gave up on after 10 minutes, I would likely have continued on with the book.

In summary, this is better than bad human narration. An average narrator is still going to beat this. It's likely you or I would do a better job than Launchpod if we read lines multiple times and got rid of all of our stumbles, but that's going to take a lot longer than copying and pasting text into this.

So with that said I'm going to give it 4 out of 5 since I've seen meaningful progress in the past month.

To the developers, keep up the good work.

The feature for creating outlines with multiple creators is certainly a strong point, offering great creative potential. However, I've encountered issues with the voice cloning feature, as I haven't been able to replicate my own voice accurately. There's room for improvement, especially in multilingual voice cloning, which currently falls short of expectations. Additionally, the music library is somewhat limited. I hope the roadmap is followed closely, and I wish for the success of Launchpod.

I was interested in the cloning aspects of this. I use Elevenlabs and their professional cloning is very awesome so I was wondering how good LaunchPod's cloning is. I uploaded 2 different voices to check out their cloning and I have to say it's very poor compared to Elevenlabs. It doesn't really sound that much like my voice at all. Elevenlabs seems so far ahead in this area that I'm not sure if other could catch up so quickly.

However, I thought the idea of creating a podcast and having a few people discussing something was very interesting - all done by AI and prompt engineering and then put nicely into a podcast through this program. So I checked out this function - it's definitely something very unique. Before I bought and checked out this product, I mentioned in a question that if it really worked, that would be a very interesting use of this product. And I thought it would involve a lot of good prompt engineering for it to work well.

So anyway, I checked out this function of this product and even though this function wasn't what made me initially interested in it, my verdict is that it's pretty good.

I created 2 podcasts with 3 people each. Selected the voices. One was 10 minutes, another 15 minutes. On a topic I'm very familiar with.

Here are some ways this aspect of the product could include:
- There were some occasions in the podcast that didn't really make sense. Whatever prompt is being used, definitely, it's not perfect, but still pretty good
- If further additional prompting could be allowed, that would be good. For example, after we review the script and want certain things done that wasn't done, instead of editing it yourself manually, further prompting could do a quicker job. Everybody's niche and expectation is different so no default prompting would be perfect. So allowing additional prompting based on the result given would be good.
- I think it's still quite obvious that this is created by AI because it lacks the conversational aspects of a podcast - the quick back and forth, interruptions, casualness sometimes and jokes, etc. If that could be added, that would be good.
- The could be further functions like creating a summary of the podcast and even a summary with timestamp - many podcasts have that these days.

Overall, I like this product. I accidentally bought 3 tiers (not sure why that happened), and I'm still not sure if I'm going to keep it. But it's definitely a unique product. I would use it purely for AI content/podcast creation if I do keep it. But I would like to see more improvements.

But I like the owner's responses and the desire to create the best product possible. Keep up the hard work.

I purchased LaunchPod AI in early July with the expectations that I could use it to record my online courses and convert my written content into audiobooks with my own voice. I created the first clone of my voice and tested it with a short paragraph while I waited for the audiobook process to be released. With the release of the audiobook capability this week I was anxious to give it a try.
I recorded the first section of the first chapter of my book. It was 430 words in length. I have a few observations after this first test:
1. The quality of the voice cloning is good, but not as good as another similar product.
2. There does not appear to be the ability to interject pauses where needed, especially between paragraphs. Perhaps that is on the roadmap?
3. The recording actually missed a couple of words, which in my mind is a material issue.
4. The actual process is a little "clunky", Not as intuitive as I would expect. But not a deal breaker for me.
5. The biggest issue for me, was the limited amount of conversion from text to speech using my voice. I only purchased Tier 1 (15 credits) to try it out with the intention of upgrading to a level I needed. But what I found with this first go was to record the first section, 430 words, using my voice consumed 8 credits. This particular book is 26,000 words. A quick estimate indicated it would take Tier 5 to record the main part of the book and I would still be a little short. This is an issue for me.
In summary, the basic product is functionally okay and maybe more suitable for blogs. For my use in producing the audio portion of online courses and creating audiobooks it is just not there yet for me.
I hope this provides the developers with some worthwhile feedback.