Show HN: Sohri – Turn short stories into binge-able audio episodes

6 jrlee 2 7/13/2025, 12:00:44 PM sohri.ai ↗
Hey HN, I love creating, and I've been fascinated by AI audio tools. But as a creator, trying to actually use them felt like a trap.

First, wrestling with raw TTS APIs to create something that doesn't sound like a robot reading a phone book is an art form in itself. It takes forever to get the pacing and tone right.

Second, and this was the real deal-breaker: even if you create a masterpiece, platforms like YouTube feel like they're just waiting to demonetize you or slap a restriction on your work because it's "AI-generated." It feels like you're building on someone else's land, and they can change the rules overnight. It's soul-crushing.

I wanted a sanctuary for audio creators. A place that’s more like a recording studio in your pocket than a frustrating API wrapper. A platform where you can focus on your story, not on fighting the tools or the distribution algorithm.

So, I built Sohri (https://sohri.ai). Sohri is designed to turn your short stories into polished, binge-able audio episodes, handling the boring parts so you can be the director. During testing, I was surprised to find that for many shorter pieces, the AI narration actually felt more engaging than my own quick recordings because it never gets tired or stumbles.

It’s free to use, and the core features will always be free. My goal is to build a place where creators can experiment, share their work, and eventually earn a living without fear. I’m here all day to answer any questions. I'd be incredibly grateful for your feedback—try listening to a story or creating your own and let me know what you think.

Comments (2)

ghjhome · 22m ago
I've been playing with the raw ElevenLabs and OpenAI TTS APIs for a while now, and the latency and quality profile here feel familiar. Can you be transparent about what your actual tech stack is? What's the "secret sauce" or the proprietary tech here beyond a nice UI wrapper, and why should I use this instead of just hitting the APIs directly with my own scripts?
jrlee · 21m ago
Great question, and this gets to the heart of what makes us different. While we utilize some external APIs for supplementary functions, our core TTS engine is our own proprietary technology that we run on our own servers. Our team (Humelo) has been focused solely on audio AI and TTS in Korea for the last 8 years. We've built a strong technology base there, but the market for audio content is small. We launched Sohri to bring our tech to creators in the larger US market and would love your support.