This plugin provides local, privacy-focused AI assistance within Obsidian, enabling offline GPT-powered text and image processing. It targets Obsidian users seeking to leverage AI for note-taking, summarization, and content generation without relying on cloud services.

How It Works

The plugin integrates with local LLM providers like Ollama, allowing users to select text or images and apply AI actions via a context menu. It supports Retrieval Augmented Generation (RAG) for incorporating context from links, backlinks, and PDF files, enhancing AI responses with user-specific data.

Quick Start & Requirements

• Install via Obsidian community plugin store or BRAT.
• Requires the AI Providers plugin.
• Install an embedding model via Ollama (e.g., nomic-embed-text for English, bge-m3 for other languages).
• Configure AI providers and Obsidian hotkeys.
• Official Docs: https://obsidian.md/plugins?id=local-gpt
• AI Providers Plugin: https://obsidian.md/plugins?id=ai-providers

Highlighted Details

• Supports text and image processing.
• Enables RAG with links, backlinks, and PDFs.
• Offers customizable AI actions and system prompts.
• Designed for maximum privacy and offline access.

Maintenance & Community

The project is developed by pfrankov, who also maintains other Obsidian plugins like Colored Tags and Obsidian AI Providers. Community interaction and shared actions are encouraged.

Licensing & Compatibility

The plugin is available through the Obsidian community plugin store, implying compatibility with Obsidian's plugin ecosystem. Licensing details are not explicitly stated in the provided text.

Limitations & Caveats

The README notes "No speedup," indicating potential performance limitations. Specific hardware and model configurations may impact usability.