reddit · SaaS · 09-19 02:50 44 分 ← 返回列表

I got tired of wiring YouTube transcripts into my AI agents, so I built this

I've been building TubeLens because I kept running into the same annoying problem: giving AI agents useful YouTube data usually means wiring together APIs, transcript extraction, scraping, etc. So I built a small API + MCP that lets an agent access things like: YouTube search → transcripts → comments → thumbnails → channel/video data The idea is basically: AI agent → TubeLens → YouTube I'm trying to validate the idea before going too far with it. Would something like this actually

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：AI agent 开发者需要把 YouTube 搜索、字幕、评论、缩略图和频道数据接入自动化流程，但手动拼接 API、字幕提取和抓取很繁琐。

目标用户：构建 AI agent、自动化工作流和 MCP 集成的独立开发者与小团队。

证据：I got tired of wiring YouTube transcripts into my AI agents, so I built this. I'm trying to validate the idea before going too far with it. Would something like this actually be useful in your automation/agent workflows?

变现：按 API 调用量订阅或预付额度收费，提供免费额度；后续可做 MCP 托管、团队协作和高级抓取/缓存/去重功能。

MVP：2 周内做一个 REST API + MCP server，支持传入 YouTube URL/关键词，返回字幕、评论、元数据，接入 Stripe 按量计费，并提供 Cursor、LangChain、n8n 示例。

竞品：Supadata、Apify YouTube Scraper、RapidAPI YouTube Transcript API、yt-dlp、youtube-transcript-api、已有 YouTube MCP servers

摘要：一名开发者因厌倦为 AI agent 手动拼接 YouTube API、字幕提取和抓取，构建了 TubeLens API+MCP，提供搜索、转写、评论、缩略图和频道/视频数据。帖子正在 Reddit 上验证需求，但没有付费意愿原话或收入数据，且 YouTube 数据 API/MCP 赛道已有多个竞品。