Team AI for Mac

Cloud & on-device LLMs, agents, and MCP tools in one native workspace — Read Me

Connect cutting-edge cloud LLMs, fully offline local models, and custom MCP tools in a single, native macOS workspace. Chat, run multi-agent roundtables, generate images, and build a private on-device memory and document knowledge base — with your keys in your Mac's Keychain and your data stored on your Mac (and mirrored to your own private iCloud while you are signed in to iCloud). Team AI for Mac is a one-time purchase and works without an account. Team AI Plus is an optional subscription (monthly or yearly, paid on a Stripe checkout page in your browser; Team AI never sees your card) that adds encrypted sync across your devices, expiring encrypted share links, Team AI Included (replies without your own API key, within a monthly allowance; it answers through Team AI’s server and Google’s Gemini API) and always-on agents (scheduled jobs that run on Team AI’s server; the prompt of a job is readable by the server and is not end-to-end encrypted). Plus needs a Team AI account (on the Mac: Google or an emailed link). Manage or cancel with Manage subscription in Settings › Team AI Plus.

Requirements

• macOS 14 (Sonoma) or later.
• Apple Silicon (M-series) recommended. The app is a universal binary, but it has been built and tested on Apple Silicon only: Intel Macs are untested. Cloud providers (bring-your-own-key) need only an internet connection. Local language models run on llama.cpp (GGUF models, Metal acceleration); Apple's MLX framework is used only by the on-device image tools. The in-app model list shows a memory check for your Mac before you download a model.
• An internet connection is needed for cloud providers, downloading local models, activating your licence and checking for updates — once a local model is downloaded, it runs fully offline.

Installation

• Download TeamAI-Mac-2.0.0.dmg from the App Market.
• Open the .dmg and drag Team AI into your Applications folder.
• Launch it from Applications.

The app is notarized by Apple, so it opens normally — there's no "unidentified developer" warning and no Gatekeeper workaround needed.

Getting Started

• Team AI includes a 14-day free trial. After the trial, a one-time purchase unlocks the app permanently — enter your license key in Settings › License. Once activated, the license is verified entirely on your Mac and keeps working offline.
• On first launch, either add a provider API key in Settings › Providers (OpenAI, Anthropic, Google, OpenRouter, and more — you bring your own key), or open Settings › Local Models to download an on-device model and use the app with no API key at all.

Features Guide

Menu Bar Mini-Chat

Click the menu bar icon for a compact chat popover with its own input box — send a quick message without opening the main window, jump into the full app, or start a new chat.

Chats

Talk to any configured model with streaming replies, markdown and code rendering, attachments, voice dictation, and a per-message action bar (copy, share, regenerate, remember, bookmark).

Agents

Create custom agents with their own system prompt, model, and personality.

Group Discussions (Roundtable)

Put multiple agents in a room to debate a topic, with an orchestrator, optional consensus tracking, voting, and a Listen mode that reads the discussion aloud with a distinct voice per agent.

Image Studio

Generate and edit images with cloud or on-device models, guide results with an existing image (image-to-image), and apply chained Image Recipes.

Memory & Knowledge Bases

A private, on-device memory bank remembers key facts across chats, with a weekly "What I've learned about you" digest, and resurfaces revisited conversations by recency. Add PDFs, text, and Markdown files as a Knowledge Base, embedded on your Mac by default and searched on-device, then attach a base to any chat to ground answers in your own documents. Choose Memory's embedding provider in Settings › Memory — the built-in on-device model, OpenAI, or any OpenAI-compatible endpoint you've configured under Settings › Providers (including a local Ollama server).

Standing Agents

Agents that run on a recurring schedule and leave their results ready for you.

Local Models & MCP Tools

Download and run offline LLMs and image models directly on your Mac, and connect Model Context Protocol servers from a built-in directory (or add your own) to give models real tools.

Usage & Costs

Track token usage and estimated spend per provider and model, with an optional monthly budget.

Privacy & Security

• Local-first. Chats, agents, memories, and knowledge bases are stored on your Mac. On-device models, memory, and document search run entirely offline.
• Bring your own key. Cloud provider API keys are yours; requests go directly from your Mac to the provider you chose — Team AI has no server in the middle of your AI requests. Anything you send to a provider (your message, attached files, and any memories or knowledge-base passages included with it) is subject to that provider's own terms.
• iCloud sync is on by default. When you are signed in to iCloud, your chats, agents, memories, MCP server names and addresses, and the text of knowledge bases, watched folders and meeting transcripts are mirrored to your private iCloud (CloudKit) so your other devices can read them. Team AI does not add encryption beyond Apple's own protection and keeps no copy of it. This version has no switch to turn it off. Provider API keys stay in your Mac's Keychain and are not synced.
• No analytics. Team AI contains no analytics, tracking or advertising SDK.
• What our licence server sees. Only licence activation and Manage Devices: your licence key, a random device id and your Mac's name. Update checks fetch the update feed from toinmoi-ai.com.
• Your license key is stored securely in the macOS Keychain and verified locally.
• Auto-updates are delivered securely via Sparkle, with cryptographically signed updates.

Support

Questions, bugs, or feedback: support@toinmoi-ai.com