Which model do you use ? my Mac is a M2 Max 32 gb ram.
Use case :
I want to create a database with my journal entries from obsidian ( all in markdown) and then ask questions related to the journal entries.
1 Like
For pure text documents I would recommend GPT-OSS:20b as it‘s relatively fast. Other useful options are IMHO Gemma 3, Mistral Small 3.x or Qwen 3 and these support also vision (e.g. images, videos or PDF documents without a text layer).
1 Like