Toi 'n' Moi® Team AI is the ultimate multi-agent AI workspace for developers, tinkerers, and power users. Chat with the world's most powerful LLMs, orchestrate teams of custom AI agents, compare models side-by-side, generate and edit images, and unlock complete offline privacy through on-device inference — all in a beautiful, native experience on iPhone, iPad, and Mac.
**RUN POWERFUL AI ON-DEVICE AND OFFLINE**
Say goodbye to API limits and privacy concerns. Download and run real, desktop-class open-source models directly on your iPhone, iPad, and Mac.
- Massive Built-in Catalog: 55+ curated, optimized GGUF models, from 230 MB edge models to 30B-class models for Mac.
- Elite Models: Qwen 3.8, Gemma 4, Llama 3.1, DeepSeek R1, Granite 4.1, Nemotron 3.5, Mistral, Phi-4, and more.
- 100% Private and Free: Chats never leave your device. Zero API costs, zero tracking, fully offline.
**MULTI-AGENT ORCHESTRATION**
Create specialized agents and watch them debate, collaborate, and review each other's work in real-time group chats.
- Instant Team Presets: 16 pre-configured expert agents. Launch a Startup Team, Code Review Panel, or Debate Club.
- Custom Agent Builder: Custom system prompts, avatars, and model backends per agent.
- Discussion Playbooks: 6 structured formats — Devil's Advocate, Six Thinking Hats, Red Team vs Blue Team, SWOT Analysis, Socratic Dialogue, and more.
- Consensus Tracking and Voting: See where agents agree and disagree, then let them vote.
- Reference Documents: Attach documents for agents to analyze through their professional lens.
**AI MEMORY**
Your AI remembers across conversations. On-device vector embeddings extract and store key facts with smart deduplication.
- Memory Browser to search, pin, edit, and manage what your AI knows.
- Memory Timeline visualizing how knowledge grows over time.
- Per-agent memory scoping alongside shared cross-thread knowledge.
- Cross-Chat Knowledge surfaces related context in new conversations.
**AI IMAGE STUDIO**
Generate and edit images without leaving your workspace, using on-device models or cloud providers.
- Text-to-Image: Turn a prompt into an original image in seconds.
- Image-to-Image Variations: Attach a source image and generate AI-powered variations.
- On-Device Model Catalog: Download local diffusion models and generate images fully offline.
- Cloud Providers: Connect Gemini or OpenAI image generation for max quality and speed.
**MODEL COMPARISON**
Compare two AI models head-to-head — send a prompt, see both responses side-by-side with speed metrics, and pick a winner.
**CONVERSATION BRANCHING**
Long-press any message to fork the conversation into an alternate timeline.
**PROMPT LIBRARY AND WORKFLOW PIPELINES**
Save your best prompts for reuse. Build sequential AI pipelines where each step's output feeds the next — automate extraction, analysis, and summarization.
**CODE EXECUTION SANDBOX**
AI writes and executes code directly in your chat, with results rendered inline.
**MODEL CONTEXT PROTOCOL (MCP)**
Connect your AI to external datasets, local filesystems via stdio, and live web data — complete control over your LLM's context and capabilities.
**CLOUD AND MULTI-MODAL POWER**
- Access Any Model: OpenAI, Anthropic, Gemini, OpenRouter, Groq, and custom providers.
- Connect Local Servers: Plug into your desktop Ollama instance over Wi-Fi.
- Multi-Modal Chat: Upload images, code files, and documents.
- Usage Dashboard: Track cloud API costs with real-time token tracking and provider breakdowns.
Whether debugging code, running models offline, generating images, or building custom agentic pipelines, Team AI is everything you need in one powerful app.
Support URL: https://www.toinmoi-ai.com/contact
Privacy Policy URL: https://www.toinmoi-ai.com/privacy-policy
License Agreement: http://www.apple.com/legal/itunes/appstore/dev/stdeula
Show more
Show less
