Local LLM Autocomplete

Tab autocomplete has been a killer feature of GitHub Copilot and similar coding LLMs for a little while now. However, this is no longer the case! The folks over at Continue launched an experimental tab autocomplete feature last week, and it’s been working great for me so far. The steps to install and use are also drop-dead simple:

• Install the Continue VSCode extension.
• Install Ollama.
• Install a model with Ollama (ollama run deepseek-coder:1.3b-base).
• You’re done (though VSCode might need a restart).

Docs for Continue tab autocomplete can be found here.

You can also install local LLMs for use in Continue’s chat using Ollama. I currently use mistral-7b on my M1 MacBook Pro and it works like a charm.