Creating ollama plugin (copilot/chatgpt like extension)

[alejandro-alzate]

Check for existing issues

• [X] Completed

Describe the feature

Cut and Dry: Forget using copilot or chatgpt with their subscriptions and privacy nightmares.

We have generative transformers at home generative transformers at home: $ ollama run codellama llama2 etc.

So interfacing with a local running one although with more latency which at least it only costs the electricity bill and a bit of heat.

So having in-house code completion hints would be nice.

If applicable, add mockups / screenshots to help present your vision of the feature

No response

[JosephTLyons]

Hey @alejandro-alzate, we currently have support for Ollama in the assistant panel:

https://zed.dev/docs/assistant-panel#using-ollama-on-macos

We also have an issue up for someone wanting to potentially add Ollama support for inline completions (copilot/supermaven style completions):

• https://github.com/zed-industries/zed/issues/14253

Feel free to upvote that with a 👍 one, I'm going to close this one out.

[alejandro-alzate]

Sorry for bothering though. but thanks!

[brianevanmiller]

Hey @alejandro-alzate, we currently have support for Ollama in the assistant panel:

https://zed.dev/docs/assistant-panel#using-ollama-on-macos

We also have an issue up for someone wanting to potentially add Ollama support for inline completions (copilot/supermaven style completions):

• Creating ollama plugin (copilot/chatgpt like extension) #14253

Feel free to upvote that with a 👍 one, I'm going to close this one out.

@JosephTLyons - it looks like you linked the same issue here.

I personally think local LLM support for inline completions would be excellent. I work on a few private codebases and don't trust Supermaven / Copilot with those codebases.

Should we open a new issue or did you mean to refer to an existing issue?