eBook to audiobook narration with realistic AI voices
For a while I've wanted to try out the new AI voices for long-form narration, but everything I found required a subscription that didn't justify my limited usage. I came across the open Kokoro model [0] and the voices are very good -- good enough to listen to for hours without the fatigue I got from legacy, robotic TTS voices. The model is 82m parameters and designed to run fast, but I still struggled to get reasonable times from CPU inference on my 12-core laptop. I thought a cloud-based GPU service would let me generate audiobooks fast enough to feed my own self-hosted library, and that…
What it does
In the maker’s words, at launch
For a while I've wanted to try out the new AI voices for long-form narration, but everything I found required a subscription that didn't justify my limited usage. I came across the open Kokoro model [0] and the voices are very good -- good enough to listen to for hours without the fatigue I got from legacy, robotic TTS voices. The model is 82m parameters and designed to run fast, but I still struggled to get reasonable times from CPU inference on my 12-core laptop. I thought a cloud-based GPU service would let me generate audiobooks fast enough to feed my own self-hosted library, and that same pipeline could become a product other people could use. I had two goals in building this: get some exposure to AI multi-agent coding workflows, and build a TTS product targeting ebook to audiobook conversion specifically. 99% of ebookaloud was written by DeepSeek v4 in OpenCode. I've used about 750 million tokens costing $12 in credits over the course of a month, and I'm very pleased with the results. Every change/feature went through a plan -> implement -> test -> review -> correct -> commit cycle with a mix of Pro and Flash agents. This was generally limited to one or two concurrent workers. I had a separate eval agent for quality control on various parts of the extraction and synthesis pipeline, which I could run 8-10 at a time. I may be approaching Yegge's Stage 6 [1] in terms of AI workflow automation. I later set up Claude Code and ran Opus 4.8 side by side with DeepSeek. There are definitely quality differences, but I'm an experienced developer with a hands-on approach. I didn't write any of the code, but I have read critical sections of what it generated and had extensive conversations with DS Pro about each step of the approach. Opus didn't have much critical to say about DeepSeek's choices, and I'm not convinced a frontier model would have made an appreciable difference for my workflow. I suspect on a large codebase the differences would become more apparent, but the few changes I implemented with Opus had similar issues to all the models I've used: random changes without my direction, over-complicating simple solutions, taking unanticipated/destructive actions when it gets stuck, etc. I do see Opus being capable of handling more of the complex planning and orchestration that I was involved in. That's something I may want sometimes but not others. As to the product itself, there are a lot more sophisticated solutions out there. I'm not trying to compete with ElevenLabs. I'm targeting m4b generation for a seamless audiobook experience with a pay-as-you-go pricing model and good-enough output quality. This is the first product I've ever tried to commercialize, and AI code generation put something polished within reach. Without AI, this would have taken me 6-8 months of manual research and development, and I would have gotten burned out long before completing it. I have a free sample on the front page of the site if you just want to see what it generates in terms of voice/format. I made a few opinionated decisions regarding output quality. I aimed for 140 wpm for most of the voices to match industry standards, but some are much slower or faster and lose prosody at that rate. Rather than give users a wall of options, I'm deferring to the playback device for things like speed control. If the site sees real usage I'd like to expand to support Kokoro's other languages, and extraction and synthesis from PDF would round out the product quite a bit. [0] https://github.com/hexgrad/kokoro [1] https://steve-yegge.medium.com/welcome-to-gas-town-4f25ee16d...
Does the same job
all alternatives →
- AAAI-Audiobook-Maker: Helping My Disabled Little Cousin Enjoy Story Books2023 · github.com · ▲5
Hey everyone! I've been working on a 'program"' (read as: overengineered bash script) It's called AI-Audiobook-Maker, and it was born out of my experiences with my 10-year-old cousin who is developmentally disabled. Despite his difficulties with reading, his enthusiasm for stories goes hard, even if he can't READ them. Watching him, I realized that the world of books was kinda out of reach for him, and I suppose many others. So, I set out to build something that could bridge this gap. After only finding shitty, expensive subscription sites that only gave you a couple hours of audio gen, I…
- CCCallhackernews.com – The missing voice interface for Hacker News2024 · callhackernews.com · ▲9
Hey there HN! I believe the future of AI communication will be more voice and less text. Low-latency realistic voice interactions are finally becoming feasible. I've built a few voice-first apps on Retell AI using Elevenlabs voices. This one uses Claude Haiku for responses and Mixtral to switch between posts and comments. The AI knows about the top 30 posts and their comments on Hacker News right now. After a Google sign-in you can try it free for 10 minutes. I'd love to hear your thoughts!
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, June 2026
the whole month →
Fundraisly▲1,544AI fundraising agent that finds investors and books meetings
AI · Jun 2026 · fundraisly.com
- H6Homebrew 6.0.0▲1,481
Today, I’m proud to announce Homebrew 6.0.0. The most significant changes since 5.1.0 are a new tap trust security mechanism, the new faster, smaller, default internal Homebrew JSON API, sandboxing on Linux, better defaults informed by our user survey, many brew bundle improvements, improved performance and initial support for macOS 27 (Golden Gate). Happy to discuss any questions here!
Dev tools · Jun 2026 · brew.sh
- PU
hope you enjoy
Life & fun · Jun 2026 · vorpus.github.io
- IM
Life & fun · Jun 2026 · hackernewstrends.com