FastCodeR1

Locally Hosted Coding Assistant!

Just install the extension (& ensure the Requirements are satisfies) and run the Chat with DeepSeek! command in your command palette.

Features

• Talk with the State of the Art DeepSeek-R1 model locally
• Have complete conversations with user chat history support
• Get fast help with your projects

Requirements

Ensure ollama is installed with the 1.5b model installed.

Run the following command in powershell after installing ollama to ensure the weights are downloaded.

ollama run deepseek-r1:1.5b  

Release Notes

0.1

This is the prerelease, it implements the idea with a basic UI. The current version supports:

• Ability to chat locally in VSCode Environment
• Direct support for 1.5b QWEN distill

The next version will have the following features:

• Ability to Change Model
• Code Block Rendering
• Think Tags Rendering
• User Chat History

Enjoy!