Comment by Adityav369

You can run this fully locally using Ollama for inference, although you'll need larger models and a beefy machine for great results. On my end llama 3.2 8B does a good job on technical docs, but bigger the better lol.