📡 Idea Radar 每天 8:00 / 18:00 推送 · 18:00 生成精选
reddit · SideProject · 51 分 ← 返回列表

Solo dev here: I spent 8 months building a PDF-to-audiobook app, and I even run my own GPU server for the voices

I'm Emin, a solo developer from Turkey. For the past 8 months I've been building Obook , an app that takes any PDF or ebook and turns it into an audiobook you can actually enjoy listening to. Why I built it: I have way more books and PDFs than time to read them. Existing TTS apps either sounded robotic, choked on messy PDFs (headers, page numbers, and footnotes read out loud mid-sentence), or cost a fortune. What it does: Import a PDF and it gets cleaned up and spli

五维评分

付费意愿
55
需求真实度
60
个人可落地
45
竞争空间
35
新鲜度
45

AI 分析

痛点:现有TTS应用声音机械、无法清理PDF页眉页脚脚注,且长书按字符付费成本很高。

目标用户:有大量PDF/电子书但没时间阅读、需要听书的学生、研究者、专业人士和重度阅读者。

证据:Existing TTS apps either sounded robotic, choked on messy PDFs (headers, page numbers, and footnotes read out loud mid-sentence), or cost a fortune. / Instead of paying per character to a TTS API, I run the voices on my own GPU server, which is the only way the economics work for long books. / Live on iOS and Android, with about 100 users so far.

变现:免费试听 + 订阅制或按听书时长/字符积分付费;高级语音、批量PDF、学术/教育套餐;也可面向出版社或教育机构做B2B API。

MVP:2周内做Web端MVP:上传PDF/EPUB后自动清洗页眉页脚脚注并按章节切分,调用现成TTS API生成音频,先不自建GPU,按字符/分钟收费验证付费转化。

竞品:Speechify、NaturalReader、ElevenLabs Reader、Audible、Google Play Books TTS、Obook

摘要:土耳其独立开发者用8个月做出Obook,把PDF/电子书转成有声书,解决现有TTS机械、PDF脏数据乱读和长书成本高的问题。产品已上线iOS/Android,约100用户,并自建GPU服务器以压低按字符TTS成本。帖子未披露收入或付费用户,但痛点真实且已有竞品验证付费市场。