Update v3.mdx

Author: mythz

content/docs/v3.mdx

@@ -10,6 +10,7 @@ description: Major release focused on extensibility, expanded provider support,
 | **530+ Models** | Access over 530 models from 23 providers via [models.dev](https://models.dev) integration |
 | **Model Selector** | Redesigned full-featured dialog with search, filtering, sorting, and favorites |
 | **Extensions** | Add features, providers, and customize the UI with a flexible plugin architecture |
+| **Gemini RAG** | Manage Gemini File Search Stores and manage document uploads for RAG workflows |
 | **Tool Support** | First-class Python function calling for LLM interactions with your local environment |
 | **KaTeX Math Typesetting** | Support for beautiful rendering of LaTeX math expressions |
 | **Calculator UI** | Beautiful UX Friendly UI to evaluate python math expressions |
@@ -19,6 +20,7 @@ description: Major release focused on extensibility, expanded provider support,
 | **Media Gallery** | Beautiful UI to browse generated images and audio generations |
 | **SQLite Storage** | Migrated IndexedDB to server SQLite for robust persistence and concurrent usage |
 | **Asset Caching** | Persistent image/file file caching with metadata |
+| **Gemini RAG Extension** | Manage Gemini File Search Stores for RAG workflows with document uploads and sync |
 
 ---
 
@@ -27,6 +29,7 @@ description: Major release focused on extensibility, expanded provider support,
 - [New Model Selector UI](#new-model-selector-ui)
 - [Rewritten for Extensibility](#rewritten-for-extensibility)
 - [Extensions System](#extensions-system)
+- [Gemini RAG Extension](#gemini-rag-extension)
 - [Tool Support](#tool-support)
 - [Core Tools](#core-tools)
 - [Calculator UI](#calculator-ui)
@@ -287,6 +290,55 @@ Smart generation models like Nano Banana's **gemini-2.5-flash-image** perform ex
 
 ---
 
+## Gemini RAG Extension
+
+The [gemini](https://github.com/llmspy/gemini) extension provides a complete solution for managing Google Gemini's [File Search Stores](https://ai.google.dev/api/file-search), enabling **RAG (Retrieval Augmented Generation)** workflows with automatic document uploads, category organization, and bidirectional sync between your local database and Gemini's cloud storage.
+
+Build up your own knowledge base in File Stores, optionally organized into categories, that you can query to ground your AI chats with your own data - whether that's searching across a single document, a category of related documents, or your entire filestore.
+
+### Install
+
+Install the **gemini** extension via the CLI:
+
+```bash
+llms --add gemini
+```
+
+After which you'll be able to click the **Gemini Icon** to open the Gemini extension page from the sidebar to manage your filestores.
+
+<Screenshot src="/img/gemini-filestores.webp" />
+
+### Key Features
+
+- **Filestore Management**: Create and manage isolated stores of documents for different projects or knowledge bases
+- **Drag & Drop Uploads**: Easily upload documents (PDF, Text, Markdown, etc.) by dragging them into the UI
+- **Smart Categorization**: Organize documents into categories (folders) for granular retrieval
+- **Contextual RAG Chat**:
+  - **Ask Filestore**: Chat with the entire knowledge base of a filestore
+  - **Ask Category**: Focus your chat on a specific category within a filestore
+  - **Ask Document**: Chat with a single specific document
+- **Bi-Directional Sync**: Reconcile your local database with the remote Gemini File API
+
+### Uploading Documents
+
+Documents can be uploaded by dragging and dropping files onto the upload zone or clicking to open the file picker. You can organize uploads into category folders by typing a category name before uploading.
+
+<Screenshot src="/img/gemini-filestores-upload-folder.webp" />
+
+Uploads are processed asynchronously by a **Background Worker** utilizing a **DB Queue**, so you can continue working while documents are indexed. The worker automatically starts when new documents are uploaded and efficiently handles batch processing without blocking the UI.
+
+### RAG Chat in Action
+
+Once documents are uploaded, you can start contextual RAG chat sessions with your data. Each session is pre-configured with a Gemini Model and the `file_search` tool to query your selected filestore, category, or document:
+
+<Screenshot src="/img/gemini-search-document.webp" />
+
+The grounded sources used to answer your query are displayed at the bottom of each chat response, allowing you to verify and explore the source documents.
+
+See the [Gemini Extension](/docs/extensions/gemini) docs for complete usage instructions.
+
+---
+
 ## Tool Support
 
 This release also includes **first-class support for Python function calling (Tools)**, allowing LLMs to interact with your local environment and custom functionality.