Have you tried NotebookLM for something like this?
2. Useless with large context. Ignores, forgets, etc.
3. Terrible code and code understanding.
Also this is me hoping it would be good and looking at it with rose tinted glasses because I could use cloud credits to run it and save money.
I’ve used it extensively to cross-reference and analyse academic papers, and the performance has been excellent so far. While this is just my personal experience (YMMV), it’s far more reliable and focused than Gemini when it comes to this specific use case. I've rarely experienced a hallucination with it. But perhaps that's the way I'm using it.
I've looked into it, but as usual with LLM I feel like I'm not getting much out of it due to lack of imagination when it comes to prompting.