AiMatch

sideOS.ai

sideOS lets Mac users dictate, ask an AI agent to read the screen and operate apps, and run background tasks. It brings meeting summaries, reminders, and connected tools into a single assistant in the Mac’s notch.

Visit sideOS.ai Compare alternatives Own this tool? Claim it
sideOS.ai screenshot

Who it's for

  • Mac users who want voice or typed requests to operate apps such as Mail, Notes, Calendar, a browser, or Finder.
  • People who need on-device dictation or meeting transcription and want those workflows to run on their Mac.
  • Mac-based teams using services such as Gmail, Calendar, Notion, Slack, or GitHub who want to connect their tools through MCP.

Not the right fit if…

  • People using Intel Macs or non-Mac systems: the page lists Apple Silicon and macOS 14 or later as requirements.

How it fits your workflow

Invoke the assistant with the keyboard shortcut, speak or type a request, and optionally point to a screen region. sideOS reads the interface and can open apps, click, type, or use AppleScript; it reports the result aloud.

For longer jobs, send work to a background task that can search the web and work with files. For meetings, sideOS offers to record when it detects a supported call window, then transcribes on-device and produces a summary and action items.

Connect listed services or other MCP servers to add tools. Dictation uses macOS speech engines; agent requests use a provider key or a local model.

Pricing

Lifetime licence

$89.90 one time

  • · One Mac licence, movable to a new Mac
  • · Every future update included
  • · No subscription or expiry
  • · Agent features including screen interaction, background tasks, meetings, memory, reminders, and MCP servers

Model usage

Billed separately by the selected model provider; the page estimates roughly 2–5¢ per typical command on default models

  • · Bring a key for Claude, OpenAI, Gemini, or Grok
  • · Alternatively, run a local model with no model API charge

Dictation

Included at no extra cost

  • · Uses macOS speech engines on the Mac
  • · No API key or per-word charge

Prices checked on Oct 24, 2026 from the vendor's site. They can change; confirm before you buy.

Key features

Screen-aware agent
Reads interface elements through macOS accessibility information and can open apps, click controls, type text, use keyboard shortcuts, and run AppleScript. Users can also point to a screen region while speaking.
On-device dictation
Transcribes speech using macOS speech engines and types it at the active cursor in any app. It supports push-to-talk and hands-free modes, and the page lists 62 languages for the classic engine.
Background tasks
Runs longer jobs in a separate loop while the user continues working. The task toolbox includes web search, file reading and writing, shell commands, page reading, and PDF rendering; some actions require approval.
Meeting recording and summaries
Offers recording when a Zoom, Meet, or Teams window opens. It transcribes on-device, separates speakers, produces summaries and action items, and can turn action items into reminders.
Reminders
Creates reminders by voice or text, supports one-time and repeating schedules, and can run a task before a reminder is due so its result is ready when the reminder rings.
MCP connections
Connects services such as Gmail, Calendar, Notion, Slack, and GitHub, and supports adding other MCP servers using JSON configuration. Tools that write or change data start disabled.

Works with

  • Gmail, Calendar, Notion, Slack, and GitHub
  • MCP servers
  • Claude, OpenAI, Gemini, and Grok
  • Local MLX models

Limitations to know

  • The listed system requirements are an Apple Silicon Mac and macOS 14 or later.
  • Agent use with a cloud model requires a user-provided provider key and incurs charges from that provider; a local model is the stated alternative.
  • The licence is for one Mac, although it can be moved to a new machine.

Alternatives to consider

  • Openbroca

    Choose it if your main need is Mac dictation that turns natural speech into polished text, rather than an assistant that also operates apps and runs background tasks.

  • Spokenly

    Choose it if you want system-wide dictation across Mac, Windows, Linux, and iPhone, rather than sideOS’s Mac-only assistant with screen actions and background tasks.

Compare sideOS.ai side by side

Getting started

Apple Silicon Mac running macOS 14 or later; the page lists Accessibility, Microphone, and Speech permissions.

  1. Download the macOS app from the sideOS product page.
  2. Install it and grant the listed Accessibility, Microphone, and Speech permissions.
  3. For agent mode, choose a supported cloud provider and add its key, or configure a local model; dictation needs no API key.
  4. Use Control-Option for the agent or Control-Fn for dictation.

Frequently asked questions

Do I need an API key?

Only for the agent when using a cloud model. Dictation uses macOS speech engines and needs no key or account of its own. The page lists Anthropic, OpenAI, Google AI Studio, and xAI as key providers, and says keys are stored in the Keychain. The app purchase is separate.