Subscribe
Sign in
Jellypod — Create customizable, AI podcasts in minutes
A customizable AI podcast studio where creators produce professional-quality podcasts effortlessly. It offers lifelike AI hosts, content integration from URLs & 30+ files, precise transcript editing, & a streamlined workflow, redefining how podcasts are made.
27
Replies
@alexanderisora Hi Alexander, we definitely think humans are needed. We think more people will be involved in podcasting than ever before.
Fewer camera film developers, but way more photographers.
I absolutely loved Jellypod's mobile app and was stoked to see it transform into Jellypod Studio! I've already setup my first podcast and am starting a second one with my wife as a co-host. It was inevitable that someone created a product like this, but they've done a killer job with the ease of the UX and quality of the outputs. Super excited to see what this becomes!
@brianswichkow Thanks Brian! That means a lot coming from someone whose built such fantastic products and marketing programs. We look forward to cheering on your AI podcasts as they grow.
This is an intriguing tool! From time to time I was thinking about making my own podcast, but I never had enough time for it.
There is a nuance though. While everybody started to use LLMs for writing texts, many people can easily spot them. Autogenerated content feels much less valuable than that written by a human. In other words, autogenerated podcasts might be valued much less than traditional ones.
Another problem is that every generated conversation flows too smoothly, there is no difference in opinions. Without this tension, stories won't be too interesting.
I have an idea of how this could be fixed. Make different hosts have different opinions (positive and negative, enthusiastic and skeptical, etc). This could be defined manually or even automatically (just imagine a TV ad with a dialogue). This difference in opinions should create interesting dynamics and will make podcasts sound less autogenerated.
Another interesting direction to work on is allowing voices to overlap. First, discussions will sound more natural. Second, background music loops create an immersive atmosphere that will be useful both for stories and for discussions.
@anton_osipov All great points. On Jellypod, our speakers do overlap. You can even add interjections in the transcript editing by wrapping text in [] brackets.
But what you want from the cohosts is totally possible too. By giving them confrontational personalities and describing the episode as a debate or a confrontation, it'll make the content that you're after.
That's part of what we mean by customizable. We believe that the podcast creator should have control over every piece of the podcast, but automated with AI.
@jason_alafgani I see. The system is more flexible than I thought. I was actually wondering what personalities are for.
BTW what do you think about adding templates (e.g. a story, dispute, calm discussion, serious story with humorous touches, etc.)? It'll teach users about the wide specter of possibilities, add diversity to generated content, and make podcast creation even more fun (because the same story could be told in very different ways).
@anton_osipov That's a great idea! Adding that to our product roadmap now.
I forgot to add that you can also add this as a text source when you start the episode. But it's still a good idea to give the users options that they might not know exist.
Looks cool, the interface is very clean and well built. It might be good to have a few demos on the homepage so people know what to expect.
Also, it'd probably be good to organize the voices by language so it's not mixed up. Most people wouldn't want to sift through 80+ entries to find one that might work for them when they're searching for a particular language.
@anthony_latona Hi Anthony, that's great feedback. Typically, people want a specific language, so we figured that Search would be enough. You just type your language of choice, and it shows all the voices with that language. But a filter sounds useful too.
Typically, the walkthrough it our home page video, but we added the launch video for the launch. Here's the walkthrough. It's the first thing you see, after you sign-in to the product.
@jason_alafgani Great demo video! I should have been more clear though about what I meant by demos: Sample output from the app. Maybe a few short, general and relatable clips from generated podcasts from the app.
Back days i built something similar but a very basic Auto Podcast Generation : https://web-production-3fcf.up.r...
Afterwards i thought it has no future potential and left it abandoned , Now i am surprised to see Jellypod , My vision turned to a perfect product which i hope will be embraced by the current gen people who would love to listen to a podcast instead of reading out a text or a long annoying pdf !
Loved the AI voices and the innovation made by you :D , It will boom !
The custom AI Voices here are a real winner! Great stuff @jason_alafgani! I can see a 100 use cases for something like this already.
This is such a badass idea, Jason. Podcasting in under 5 minutes? You’ve just called out the exact reason most people don’t even start—because it’s such a freaking process. Exactly the reason why I do not have a podcast even tho I wanna talk about AI-Human relationship so badly.
What’s the next big step forJellypod? You’ve got me curious.:-)
@lenka_mezulanik Thank you for the kind words! There's so much in store. We'd like to help our creators expand into Youtube. Plus enhanced multi-language support, better natural language, more conversational style controls, broader collaboration, and faster automation are on the horizon too.
I just used your product! Overall I like it very much, and I also have such a need, but I found a big problem, it seems to often go wrong, and a new page must be refreshed.
@jasminechzi Hi Jasmine, I'm sorry for the issues you're experiencing. That's odd. Would you be willing to message support using the Blue icon in the bottom right corner? We want to get this fixed right away.
As a teacher, I'm constantly sharing knowledge and wishing I could reach more people through audio.
@jennifer_cooker That's exactly our goal. How can we help people share their story and knowledge as easy as possible. We even have college professors using at as course accompaniments to bring the content to life!
@sentry_co Haha so many voices. We have another voice featured called "Magic Voice Design". In this case, you decide what the voice's description is and the AI makes you a few options.
1000 credits is about 1000 secs of audio. We've seen some people use it all on 1 episode, and others spread it across 4 ~4min episodes.
Every "Chapter" in the Jellypod outline editor is equal to ~2-4 minutes. So you just need to adjust the default 3 chapters to 1, and away you go.
I made a 2 min, 50 sec episode about our press release, listenable here: https://jellypod.ai/blog/studio-...
@jason_alafgani Nice. Its pretty good! I experimented with googles take on this a few months back. But it lacked the customization you have. I think it could be really good for mocking podcasts as well. Like say your making an IRL podcast. Then you could conjur up some variations on how that IRL podcast could go. and then pick one that worked the best. And then not inact it 100% but you know. Build thous internal mental models prehand sort of. Anyways. Gd luck with the launch today. 🚀
@sentry_co That's an awesome idea. We've always felt like experimentation and IRL podcast augmentation is key. You can start with the AI, validate an audience and niche, and then start having longer form discussions IRL.
Jellypod