Obsidian plugins with #ollama#

The AI Image Analyser plugin sends image files to an Ollama model and stores the returned keywords inside your vault for later reuse. It works with JPG, JPEG, PNG and WebP images, and lets you trigger analysis from the context menu or command palette. The default setup uses llava-llama3, though the model can be switched to llava, llava:13b or llava:34b in settings. Results are cached in a JSON file inside the plugin folder, which means repeated analysis is avoided and cached data can be synced across devices. It also connects with OmniSearch, so image analysis results can be indexed and searched when its indexing option is enabled.

The Neural Composer plugin adds graph based AI chat to your vault by running a LightRAG server, indexing notes into a knowledge graph and combining graph traversal with vector search for more connected answers. It can start and stop the backend automatically, watch a folder for reindexing, show processing status in the file explorer and open a 2D or 3D graph view for exploration. Answers include citations linked to the source notes and text chunks, and the plugin can expose graph tools through MCP clients. It also supports remote LightRAG servers, multiple model providers and a fully local setup with Ollama, while keeping API keys in the local vault data and avoiding runtime telemetry.

The Large Language Models plugin brings chat based access to cloud and local language models into Obsidian through a sidebar, tab view and floating popup window. It supports hosted providers such as OpenAI, Anthropic, Google and Mistral, while local use can run through GPT4All or Ollama. Models can be switched from each chat surface, and the plugin includes commands to open the different interfaces or toggle the floating action button. It also adds a built in MCP server that can expose vault operations to compatible desktop clients with a local bearer token, confirmation prompts for write actions and localhost only binding. The plugin is aimed at keeping model interaction, local runtimes and vault level tool access inside the same workspace instead of relying on a separate companion bridge.

The Mini-RAG plugin enables local retrieval augmented generation by connecting your notes to a locally running LLM through Ollama. You can start a chat in the context of a specific note or folder, allowing the model to reference only relevant content when generating responses. It supports any Ollama-installed model and provides controls for model selection, temperature adjustment, and even context-free chatting when you want unconstrained responses. Interactions can be initiated directly from right-click menus in the editor or sidebar, and conversations can be saved for later reference.

The Graphene plugin adds a semantic graph that connects notes by content similarity using vector embeddings. Dotted lines show semantic relationships alongside solid wikilinks, with a D3.js physics layout and adjustable connection thresholds. Embeddings can be generated locally through Ollama or through OpenAI with an API key. Word limits, initial word skipping and heading exclusion control which text is embedded. The plugin tracks new and modified files so updates can process only changed notes. Graph controls cover forces, text fading, tags, arrows and animations. Filters handle attachments, unresolved links and orphans, while search matches file names, paths and contents. Color groups use plain text matching rather than advanced search syntax. Semantic controls include links per note and adaptive pruning. Display changes use cached embeddings without regenerating vectors.

The LLM Tagger plugin enhances note organization in Obsidian by using locally running large language models via Ollama to automatically generate relevant tags. It processes notes efficiently, avoiding re-tagging unchanged files, and can create brief summaries alongside the generated tags. Users can customize their tag list for more focused categorization and select different LLM models for processing. The plugin also supports an auto-tagging feature that applies tags to newly created or modified notes. With local processing, it ensures privacy and speed while maintaining a seamless tagging workflow.

The Tezcat plugin brings AI-powered semantic search to your notes using local or cloud-based embedding models. Inspired by the remembrance agent concept, it surfaces related content in real time-both at the note and section level-while you write. Results can be used to insert links, inject text, or jump to the source note. It supports both vector and hybrid search (combining vector similarity and full-text match) to improve relevance. Tezcat integrates best with [Ollama](https://ollama.com/download), letting you run local models privately. It also supports OpenAI embeddings for those preferring cloud-based setups. By indexing your thoughts intelligently, Tezcat enhances creative flow and contextual recall during writing.

The Ollama plugin integrates the Ollama AI tool into Obsidian, enabling users to enhance their notes with advanced text manipulation features. Pre-configured prompts such as summarization, explanation, expansion, and various rewriting styles (formal, casual, active voice, bullet points) streamline text editing and content generation. Users can also create custom prompts with specific models and temperature settings. The plugin processes selected text or entire notes and inserts the results back into the document, enhancing productivity and creativity.

The Coherence Wizard plugin combines bulk file operations with optional AI processing through Ollama. It splits notes at headings, dates or dividers, merges related notes and combines files into one document. Folder moves preserve existing destination files by adding numeric suffixes to conflicting filenames. Other tools apply frontmatter templates, reformat date filenames, copy matching content into category folders and find duplicate files for review. AI tools generate summaries, filenames, rewritten notes and frontmatter ratings. Categories can be added as tags or backlinks, with optional moves and copies. AI requests send note content and relevant prompt data to the configured Ollama endpoint, which can be local or remote. File operations work without Ollama. It requires desktop version 1.13.0 or later and does not support mobile. There is no plugin wide undo or automatic backup and failed batches can leave some operations completed.

The Vector Search plugin enhances Obsidian with semantic search capabilities using Ollama's embedding API. Unlike traditional keyword-based search, it finds notes with similar meanings by generating vector embeddings for markdown content. The plugin enables users to search for related notes based on selected text, configure similarity thresholds, and process files efficiently with automatic indexing and updates. It includes smart text chunking strategies to optimize search results and supports real-time monitoring of file changes.