Sider Omni Sidebar

Product Introduction

1. Definition: Sider Omni for Mac is a context-aware, multi-model AI agent application that operates as a persistent sidebar directly attached to any active application window. Technically, it is a macOS utility that leverages system-level permissions (Accessibility and Screen Recording) to provide real-time, in-context AI assistance without requiring manual data transfer.
2. Core Value Proposition: It exists to eliminate the disruptive workflow of copy-pasting content between work windows and separate AI chat interfaces. Its primary value is providing a persistent, per-window AI teammate that understands on-screen context, enabling seamless, focused AI interaction directly within the user's existing workflow environment.

Main Features

1. Per-Window, Docked Sidebar: The Omni sidebar is not a floating window; it docks directly to the edge of the active application window (e.g., Safari, Microsoft Word, Preview). It utilizes macOS window management APIs to maintain its position relative to the host window, moving, resizing, and even entering fullscreen mode alongside it, ensuring the AI context is always visually coupled with the relevant content.
2. Multi-Model, Agent-Based Architecture: Each Omni sidebar instance is an independent AI agent capable of selecting from multiple large language models (LLMs), including GPT-4o and other leading models. Users can switch models per-task based on performance or cost. The agent runtime operates locally on the Mac where possible, managing conversations and tool calls.
3. Contextual Screen Understanding & Tool Use: When a user query requires it, the Omni agent can invoke a "see screen" tool. With user permission, it captures and processes content from the specific, attached window (not the entire desktop) to provide context-aware actions like summarization, data extraction, or editing, all without manual file uploads.

Problems Solved

1. Pain Point: Context switching and workflow fragmentation caused by toggling between a primary work application (like a PDF reader or email client) and a separate, detached AI chatbot window or web browser tab.
2. Target Audience: Knowledge workers, researchers, students, and professionals who frequently use AI for tasks within specific documents and applications. Key personas include: Content Writers drafting in Word or Google Docs, Analysts working in Excel or Sheets, Researchers reviewing PDFs in Preview or Safari, and Executives managing email in Outlook or Apple Mail.
3. Use Cases: Summarizing a lengthy contract in a PDF viewer; drafting an email reply based on a thread visible in Mail; extracting action items from meeting notes in Obsidian; generating slide content from a bullet list in PowerPoint; researching a topic by analyzing the webpage currently open in Chrome.

Unique Advantages

1. Differentiation: Unlike ChatGPT Desktop or Claude.app, which are standalone applications, Omni is a window-attached utility. Unlike Microsoft Copilot, which is limited to Microsoft 365 apps, Omni is application-agnostic, working with any macOS window. It provides persistent, per-window context where competitors offer either global screen analysis or no persistent context at all.
2. Key Innovation: The "one agent per window" model is its core innovation. This allows for multiple, simultaneous, and isolated AI conversations tailored to different tasks (e.g., one for a spreadsheet analysis and another for writing in a document), with the optional ability for a single agent to perform cross-window tasks upon user request, a flexibility unmatched by other desktop AI assistants.

Frequently Asked Questions (FAQ)

1. Does Sider Omni work with Google Chrome and Safari? Yes, Sider Omni is designed to work with any application that creates a standard macOS window, including all major web browsers like Google Chrome, Safari, Microsoft Edge, and Mozilla Firefox. The Omni sidebar attaches directly to the browser window to assist with web-based research and content.
2. Is Sider Omni free to use, and what AI models does it support? Sider Omni operates on a credit-based system and can utilize your existing ChatGPT or Claude API credits. It supports multiple AI models, allowing you to choose between different providers (like OpenAI's GPT-4o) per task based on your needs and credit balance.
3. How does Sider Omni handle privacy and screen data? Privacy is a core design principle. Omni only accesses window content when explicitly triggered by a user query that requires the "see screen" tool, accompanied by a macOS permission prompt. Screen Recording and Accessibility permissions are separate and revocable. Conversation data is managed by a local agent runtime, and the company states it does not sell user data or use it for advertising.
4. Can I use Sider Omni in fullscreen mode on my Mac? Yes, a key feature of Sider Omni is its ability to enter fullscreen mode alongside your application window. The Omni sidebar shares the screen space in a split-view configuration, allowing for deep focus work without losing access to the AI assistant.
5. What macOS permissions are required to install Sider Omni? Sider Omni requires two standard macOS permissions: Accessibility (to control window attachment and UI elements) and Screen Recording (to capture on-screen content for contextual understanding when a tool is used). Basic chat functionality works without Screen Recording enabled.