reddit · SideProject · 10-21 02:50 48 分 ← 返回列表

I built CraveBeam, a tool that explains restaurant menus in plain words, because I never understand them

I am a picky eater, and I am tired of guessing at menus. A dish gets described in a line of chef language, I cannot picture it, so I order the same things I always order. So I built CraveBeam. You point your phone camera at a menu, or upload a photo, a PDF or a link, and it explains each dish in plain words: what it is likely to taste like, the texture, and whether it fits you, based on a quick setup where you say what you avoid. No app to install, it runs in the browser. It is free during

五维评分

付费意愿

需求真实度

个人可落地

竞争空间

新鲜度

AI 分析

痛点：看不懂餐厅菜单上的菜名或厨师术语，不知道该点什么、是否适合自己口味或饮食限制

目标用户：挑食者、有饮食限制或过敏顾虑的食客、常用手机点餐的旅行者

证据：I am a picky eater, and I am tired of guessing at menus... It is free during early access.

变现：免费增值，基础拍照解释免费，订阅解锁个性化口味档案、多语言菜单、保存历史与无广告；或按次付费/餐厅合作导流

MVP：用手机拍照或上传菜单图片，OCR 提取菜名，调用 LLM 结合用户设置的忌口生成通俗解释（口味、口感、是否合适），浏览器端即可用，无需安装

竞品：Google Lens、各类菜单翻译/点餐App、ChatGPT 拍照问答

摘要：作者是产品经理，用 Lovable 和 Claude 在业余时间做了 CraveBeam，可拍照或上传菜单，用通俗语言解释每道菜的口味、口感和是否适合自己的忌口。目前早期访问免费，无收入数据，也无他人付费意愿佐证。痛点是真实的挑食者点餐困扰，但竞争激烈、变现信号弱。