I wrote today’s post using Companion, an Obsidian plugin, that allows me to use LLMs for writing suggestions and completions.

It’s running locally using Ollama, so 100% free of charge. As I begin typing, the model starts generating text based on my input, offering suggestions and completions that aim to enhance the writing process.

It’s simple, yet quite impressive.

I saw it in action on this video and thought I had to try it.

I started using the Llama3:latest model I had installed a while ago, but I have now just downloaded the gemma2:9b model (same as in the video).

Will this be the tool that ends up making me …a better, more consistent writer?

It’s certainly promising. The ability to have nuanced suggestions and completions at my fingertips could significantly elevate my writing speed and quality, but also help at getting out of a block.