Played around with Ollama for a bit but have not tried any PDF or vision stuff yet. Would you share your setup with examples in a long form post?
Here's my drop-in replacement for ChatGPT w/ "vision" capabilities on an M2 Mac with 16GB or better.
1. Orbstack (https://orbstack.dev) +
2. Ollama (https://ollama.ai) +
3. llava model (https://ollama.ai/library/llava) +
4. ollama-webui (https://github.com/ollama-webui/ollama-webui)
Caveats: this does require a moderate degree of technical ability and comfort with the command line; ollama-webui doesn't yet support web browsing.
But with that, you've got a fast, friendly, private toolset for generative AI, including analyzing images, PDFs, and the like.
Discussion
Doubtful that I’ll have time, but I’ll see what I can do.