Current best small model for data parsing and indexing?

Lets say I'm writing my own journal/diary and want an LLM to parse everything and index names and keywords so I can ask later things about my life, which local small LLM would be a good fit that I can trust won't go...