Prosed
Automatically turn existing content like newsletters, podcasts and blogs into publish-ready nonfiction book manuscripts
In-depth Report
-
Prosed is an AI publishing service for creators that helps you automatically organize scattered content such as newsletters, podcasts, blog posts, course notes, etc. accumulated over the years into a non-fiction book manuscript that can be published directly. Its core selling point is not "writing for you", but "assembling it for you" - through a multi-stage pipeline called Inkwell, it reorganizes your existing text and voice content into a fully structured manuscript, and then cooperates with the built-in AI reviewer Typo to perform editing checks, and finally outputs print-grade PDF, ePub or DOCX. During the beta period, each book is priced at $47, and the first 100 founding users lock in the price for life.
-
Prosed is built by an independent developer named Elise (contact hello@tryprosed.com). From the perspective of product positioning, it targets a very specific pain point: many content creators publish a large amount of content on Substack, YouTube, LinkedIn, podcasts and other platforms. The content itself is already high-quality and in-depth, but it is scattered everywhere. To organize it into a book with a complete structure, the workload is no less than writing a book from scratch. The developer himself has had this experience - he almost wrote a book in 2016, but then dropped it. However, the content accumulated in the past ten years is actually enough to write a book, and the only thing missing is the "assembly" link. This market positioning is very smart. AI writing tools currently on the market can be roughly divided into two categories: one is general-purpose chat AI (ChatGPT, Claude, etc.), in which you need to conceive the framework, repeatedly prompt, and manually arrange it; the other is AI tools designed specifically for books, but most of them still use the routine of "enter a prompt word and spit out a bunch of text," and what comes out lacks the author's personal voice. Prosed chose the third path: not to generate, but to reorganize.
-
Prosed's workflow is divided into five steps. Content import. All you need to do is paste a link—Substack, YouTube channel, podcast RSS, LinkedIn, blog—or directly upload transcripts, notes, audio and video files. Prosed automatically crawls and transcribes. This step covers the vast majority of creators’ content forms, and the threshold is very low. Inkwell tubing assembly. This is the core technology of Prosed. Its multi-stage pipeline, built on patterns from more than 150 best-selling nonfiction books, will automatically reorganize your scattered content into a chapter structure and add "bridge paragraphs" where needed to connect different articles or topics. Each bridge paragraph is marked and you can modify or delete it at any time. According to official data, more than 90% of the text in the final manuscript is the author's own original text, and AI is only responsible for cohesion and structural arrangement. Private editing portal. You can browse each chapter in a clean, distraction-free reading interface, leave comments, accept or reject suggestions, and request rewrites. All editing operations are done within this portal, no need to switch tools. Typo AI review. This is a built-in editorial reviewer that checks for grammar, tone consistency, flow, chapter coherence, and brand consistency. Each review comment is linked to a specific paragraph location, allowing you to make precise revisions. Export and publish. The final output is print-grade PDF, ePub or DOCX format, which can be directly uploaded to platforms such as Amazon KDP for self-publishing. It is worth mentioning the Source Map function - it will color-code whether each piece of text is your original text or the connecting content added by AI, and display the proportion (such as "96% of your text / 4% of the connecting content"). This transparency far exceeds similar products.
-
Prosed’s pricing strategy is interesting. The beta period costs a one-time fee of $47 per book, with the first 100 founding users locking in that price, after which the standard price rises to $197. In comparison, traditional ghostwriting starts at US$15,000, plus royalties; general AI tools (ChatGPT/Claude) cost US$20 per month, but you have to be your own editor and project manager; similar AI publishing tools cost US$97, but they basically generate a text block at once. Prosed's $47 pricing is really competitive during the beta period, and it's a one-time fee with no subscription. Each book includes 5 full-text rewrites ($19 each for excess), but unlimited Typo reviews and chapter-level editing.
-
Based on the evaluation of various platforms, users’ evaluation of Prosed mainly focuses on the following points. Positive review. Users generally believe that Prosed truly solves the pain point of "a lot of content but no time to organize it into a book". A user who used Prosed wrote on Product Hunt: "When users upload enough content, the effect of Prosed is amazing. I encourage everyone to try it. $47 for a book, even if you are not satisfied, it is only the cost of a book." Another reviewer praised Prosed's design style: "The design and experience are great. This is a long-term play for the creator economy." Negative feedback. The main concerns focus on several aspects: First, Prosed is highly dependent on the quality of existing content. If the original content is logically confusing or the theme is scattered, the output effect will be compromised; second, it currently only supports non-fiction books, and novel users cannot use it for the time being; third, there are only 100 places in the beta period, and the price will more than quadruple thereafter, and the price/performance ratio will drop significantly. In addition, some comments pointed out that Prosed currently lacks a "source to chapter" mapping map, allowing authors to more clearly track which original content each paragraph of text comes from.
-
In the AI writing tool track, Prosed has a very unique positioning. The market for general-purpose AI writing tools is already quite crowded, but full-process tools like Prosed that go from “fragmented input to systematic output” are still scarce. Some Chinese media positioned it as "filling the gap from fragmented input to systematic output." From a technical and product perspective, Prosed's Inkwell pipeline is built on the analysis of the structure of more than 150 best-selling books, which is more structured than purely relying on large language models to generate content. Its Source Map feature also adds transparency, which is a plus in the world of AI writing—users are no longer faced with a black box. In terms of competing products, compared with other writing tools such as Phrasly (focusing on AI text humanization), Prosed’s core competitive barrier is not in the quality of production, but in workflow integration: integrating content ingestion, structural reorganization, editing review, and format export into a complete publishing pipeline.
-
Prosed’s biggest current uncertainty comes from its quality stability after scale expansion. There are only 100 spots available in the beta phase, so the team can focus on each project one-on-one. It remains to be verified whether the Inkwell pipeline can maintain the same output quality once opened to larger scale. In addition, Prosed's dependence on Claude API (Anthropic) is also a potential risk point. Although the privacy policy clearly states that Anthropic will not use API inputs for model training, the availability of third-party APIs, price changes, and policy adjustments may affect Prosed's operations. There is another point worth noting: Prosed currently seems to be a product of an independent developer or a very small team (the contact email is hello@tryprosed.com, there is no clear company registration information). This "packing light" model is highly flexible in the early stages of the product, but long-term product iteration and customer support capabilities need to be observed.
-
Prosed is best suited for the following categories of people. Substack or newsletter author: Over the years, I have accumulated hundreds of issues that I would like to package into a book. Podcasters and video creators: Have a lot of spoken content that you want to turn into written text. Bloggers and long-form writers: years of blog posts that need to be structured into a coherent book. Expert authors: have deep professional knowledge but no publishing experience, and need someone to help them convert content into manuscripts. Not suitable for: Creators who want to write novels from scratch; Beginners who do not have enough content accumulation; Traditional authors who pursue complete originality and are unwilling to let AI intervene in any creative process.
-
Prosed accurately addresses a real need - to transform daily information input into publishable knowledge products. It does not try to replace authors, but acts as an "assembler of content." This differentiated positioning allows it to find its own foothold in the crowded AI writing track. The $47 beta price is definitely worth a try, especially for creators who already have a content pipeline. Whether it can maintain quality after the scale is expanded and whether it will remain competitive after price increases are two key points worth paying attention to.
User Reviews
-
Raymond647—说实话，Beta 期的 47 美元定价让我觉得这产品有点太便宜了。等涨到 197 才算合理。现在先薅一波羊毛，哈哈。 -
Cheryl_Miller_20227—不知道有没有人跟我一样，我是先把内容喂进去看它怎么分章的，然后自己再调整大纲。这个自由度挺好，不是说 AI 分完就不能改了，每个章节都能自己调顺序、删减、合并。比我想象中灵活。 -
DeborahBell_lab—47 美元确实便宜，但 197 就有点犹豫了。毕竟不是每个人都有一堆现成内容可以丢进去。 -
DorisFisher_2023—PDF 导出格式不错，直接拿去 KDP 就能用。省了我很多排版的时间。 -
丁兰素—收到书稿的第一版，看了看大概有 92% 的内容是我自己的原文，8% 是 AI 加的衔接。这个比例我觉得挺合理。要是 AI 写的比例太高，那到底算我的书还是它的书？现在的做法我是认可的。 -
曹珍—反馈给团队一个问题，周末发的邮件，周一早上就回了，响应速度好评。 -
SMorris_Max—对于有内容积累的创作者来说，这就是个效率工具。不需要从零开始写，不需要手动整理稿子，不需要担心排版格式。一键导出 PDF 直接上架亚马逊。省下的时间精力远不止 47 美元。 -
happywolf620—刚开始用的时候以为它会像 ChatGPT 那样直接给我写一本书出来，结果发现它其实是个「编辑」而不是「作者」。理解这一点后体验就好很多了。如果你已经有大量内容只是没时间整理，它超好用；但如果你只有个想法想从零开始写，它帮不了你。 -
狗狗569—试了下把我的博客文章导入，确实能自动生成章节结构。但我发现它有个问题——如果文章主题跳跃太大（比如我既写编程又写旅行），它会强行串在一起，逻辑不太顺。所以它更适合那些内容一直围绕某个固定话题的创作者，东一榔头西一棒槌的就不建议了。 -
JWood_20205—Claude API 驱动的，隐私政策说不会用输入数据训练模型，这点放心。但如果哪天 Claude API 出问题了 Prosed 也受影响。独立开发者的产品，联系邮箱是 hello@tryprosed.com，没有公司实体信息，长期看迭代和支持能力存疑。 -
MsEmiliePark—作为播客主，我一直想把节目内容整理成一本书，但想到要自己转录、归类、写衔接就头大。Prosed 这个「内容到书籍」的流程确实解决了我的痛点。把 RSS 链接一贴，它自动转录加组装，门户里还能逐章修改。47 美元买一个出版级的初稿，我觉得很划算。 -
Zachary245—Typo 审阅功能挺实用，不是那种简单的语法检查，是真的会分析各章节之间语气是否一致、故事线是否连贯。而且每条意见都直接链到对应段落，改起来方便。不过第一次用的时候有点懵，建议加个引导教程。 -
SPwil—等了一年多了，终于有人做了这个。我不是作家，但写了 5 年的 LinkedIn 帖子，加起来足以出一本书。47 美元买不了吃亏，已下单。 -
Sean_Roberts_2024—最大的顾虑是价格。Beta 期 47 美元确实便宜，但名额用完后就涨到 197。如果质量能稳定还行，可 beta 才 100 个名额，团队能一对一关注，开放了还能保持吗？ -
SaraMendoza_668—试了一下，把我的播客 RSS 贴上去，AI 自动转写好章节初稿。Inkwell 管线识别得不错，把几期同样话题的节目合并到了同一章。不过播客转录确实比较乱，需要手动调不少。如果你主要做文字内容（新闻通讯、博客）效果最好，口语内容凑合能用。 -
BeityDavis—非常理解「内容很多但没时间整理成书」这个痛点。我写了六年 newsletter，每期 2000 字，算下来都够出三四本书了。Prosed 这个名字取得好，process + seed 的合体？不管了，47 块我先冲了。 -
淡然507—能写科幻小说吗？不行，目前只支持非虚构。这挺好，专注一个方向做好比什么都想做什么都做不深强。 -
MJones—我有个疑问，「听起来像你」到底是指模仿你的用词习惯，还是真的能保留你核心的观点和论证逻辑？我觉得如果只是表面语气匹配，长期来看价值有限。建议 Prosed 做个来源到章节的映射图，让人能追踪每段文字来自哪篇原始内容、哪些是 AI 加的衔接。团队说已经在路上了，期待。 -
John.Stephens—之前想过好几次要把我的 Substack 整理成一本书，但每次看到几百篇文章就头疼。Prosed 帮我一次性抓取所有内容，自动分好章节，47 美元确实值，第一版我已经收到在门户里看了，结构比我预期的好太多。最大的惊喜是那个 Source Map，能清楚看到哪些是我写的哪些是 AI 加的桥接段落，透明感拉满。不过想要出书的话最好有个 30 篇以上同主题的内容，不然效果会差点意思。 -
天涯892—当用户上传了足够多内容时，效果确实令人惊叹。47 美元换一本书，就算不满意也不过是一本书的成本。 -
Christian253—产品设计审美在线，在同品类里算是一股清流。