Plugin submission: AI Chatbot Assistant

I have been impressed by the recent progress in performance of open-source, self-hosted Large Language Models (LLMs). In the near future, the integration with offline applications like Joplin could create a promising niche for small LLM models in the 7/8B class that can run on approximately 8GB of VRAM. A swift response is crucial for daily note-taking. I found that 'Jarvis' and 'ChatGPT Search' work with online services. For the lack of time and skill, I decided to fork the latter to serve my purposes.

Indeed, this plugin helps me a lot!!!

Key improvements:

  1. Grammar: Fixed "folk" to "fork," which better fits the intended meaning.
  2. Clarity: Streamlined sentences for improved readability and logical flow.
  3. Technical terms: Ensured consistent use of technical jargon (e.g., "open-source").
  4. Word choice: Replaced vague phrasing like "the nearest future" with more precise language.

AI Chatbot Assistant

Description

Integrating Joplin with a self-hosted AI platform like Open WebUI, the plugin AI Chatbot Assistant assists you with providing a powerful AI deployment platform to operate entirely offline.

How to use

  • Go to Tools > Options > AI Chatbot Assistant Plugin Section.
  • Enter Base URL and Model
  • Restart Joplin to enable the plugin

Customization

  • Base URL: Enter your Open WebUI instance URL, or like https://chatgpt.com
  • Model: Specify a model for the chat session, like gpt-4o
  • Additional Parameters: In Open WebUI, chat sessions can be customized through various URL parameters.
    • temporary-chat: Mark the chat as a temporary session if set true.
    • web-search: Enable web search functionality if set true.

How it works

Right-click on a selected text shows the context menu with Command Names as specified in the Plugin Setting Section. Chatbot's response to the query or prompt is displayed in an external browser.

This plugin is enhancement of Joplin GPT Search Plugin by Bhashkar Kumar Gupta.

4 Likes