I hard fell in love with Notebooklm this week. I tried using it before but something clicked this week and when I realized its power it blew me away like a mushroom trip. I have 40,000 books on my computer. When I uploaded a few collections and played around with Notebooklm it had me up till 4am.
One of the videos it made is. Something else. You would have to see it to believe it.
My favorite apps right now are Claude and Notebooklm so I wanted to say thank you very much for this post 🙏
Try combining Claude agent orchestration with NotebookLM. They can continuously interact with NotebookLM within the limits you set, so you don’t have to stay in the loop.
So i have some articles with me, for which i need to create images and cover page. So asking claude to upload the articles in notebook and generating them!
Brilliant! I guess you’ll save a lot of time by automating your post covers. And they’ll be even better, since it understands your article from start to finish.
Awesome post, thanks for sharing!! I’m currently building my LLM wiki and trying to understand it better. From what I see, the difference is that what I’m building is meant to help me THINK better, whereas what you’re showing is more about BUILDING things, correct?
I hard fell in love with Notebooklm this week. I tried using it before but something clicked this week and when I realized its power it blew me away like a mushroom trip. I have 40,000 books on my computer. When I uploaded a few collections and played around with Notebooklm it had me up till 4am.
One of the videos it made is. Something else. You would have to see it to believe it.
My favorite apps right now are Claude and Notebooklm so I wanted to say thank you very much for this post 🙏
Glad you liked it, Tyler. I’ve had similar nights with NotebookLM, so I totally understand you :)
I connect Notebook LM to Claude Code using a third party Notebook LM MCP and CLI solution. It works really well!
Great job, Neill! Curious, what will your first use case be?
I tried this and able to set it up in under 5 minutes. All i had to do is authenticate.
To do this, i copied the content from your notebook, and asked claude to follow the steps and it worked!
Great method, training the Claude from this post, well done Hari! What will be your first use case?
If you have guide for usecases i am eager to look at!
Try combining Claude agent orchestration with NotebookLM. They can continuously interact with NotebookLM within the limits you set, so you don’t have to stay in the loop.
So i have some articles with me, for which i need to create images and cover page. So asking claude to upload the articles in notebook and generating them!
Brilliant! I guess you’ll save a lot of time by automating your post covers. And they’ll be even better, since it understands your article from start to finish.
Connecting Claude to NotebookLM is a clever combo. Grounded reasoning with better recall.
Thanks, Aniket. How will you use it after combining both of these tools?
Awesome post, thanks for sharing!! I’m currently building my LLM wiki and trying to understand it better. From what I see, the difference is that what I’m building is meant to help me THINK better, whereas what you’re showing is more about BUILDING things, correct?
Might give it a try as well.
For these use cases, yes. But NotebookLM is also very powerful for thinking more clearly.
You can use it as a source for anything you build.
Your system can query NotebookLM whenever it’s in doubt, which helps produce more grounded and reliable answers.
What’s the difference in connecting notebook llm to Claude vs connecting obsidian to Claude?
I believe you’re asking in terms of memory.
While NotebookLM can also be used as memory for Claude, the goal here is also to leverage its research capabilities.
In Obsidian, you need to upload documents and use them as the “brain” for your app.
But with NotebookLM, you can let it handle the research as well, and take advantage of its additional features.