Description of App
It supports GGUF, MLX and Whisper models or connect to remote AI servers, on-device MLX Audio Engine with TTS, STT, live transcription, and real-time translation.
NEW: Novel Writer - Complete AI Fiction Studio
Write full novels from a single premise. Choose from 20 genres - Fantasy, Sci-Fi, Mystery, Romance, Thriller, Horror, Historical Fiction, and more. The AI builds a structured outline then writes each chapter with genre-tuned style, keeping voice and pacing consistent throughout. A Consistency Database tracks characters, world rules, and events across chapters to eliminate plot holes. Enhanced Writing Rules (20+) cover show-don't-tell, dialogue subtext, and genre-specific guidance. A per-chapter Quality Report flags AI writing patterns. Works on a 9B on-device model. Export as EPUB.
NEW: Soul - AI Character Roleplay Studio
Turn any AI model into a fully realized character. Define name, personality, scenario, and background - Soul injects it into every chat. Import TavernAI character cards (PNG) or build from a template library. Each soul supports a custom avatar, opening message with alternate greetings, post-history instructions, and a Lorebook for keyword-triggered world-building lore. The status bar shows the character's name with a sparkle animation while generating.
Run Any Model, Anywhere
- Offline GGUF and MLX inference on iPhone, iPad, and Mac via llama.cpp
- Connect to OpenAI, Claude, Gemini, DeepSeek, or any OpenAI-compatible server
- Apple on-device Foundation Model (iOS 26+)
- Switch local and remote models mid-chat without losing context
AI Agents and Tools
- LLM Council: multiple AI models debate in parallel, chairman synthesizes a conclusion
- 40+ built-in tools: web search, news, calendar, HealthKit, code execution, and more
- Full MCP (Model Context Protocol) support with integrated Marketplace
- Siri, Shortcuts, and Share Extension integration
Voice
- MLX Audio TTS (Kokoro-82M, Fish Speech S2 Pro) - fully on-device
- OpenAI TTS (gpt-4o-tts, gpt-4o-mini-tts)
- Real-time voice chat and live transcription
Files and Content
- Convert PDFs, Office files, EPUB, YouTube, and audio to clean Markdown
- OCR, Mermaid diagram rendering, LaTeX math
- Generate images locally or via OpenRouter, gpt-image-1, FLUX.1-dev
- Export to Markdown, PDF, HTML, EPUB, or JSON
Accessible to Everyone
Sound effects and haptic feedback confirm AI events: generation, errors, clipboard, and more - giving both sighted and blind users non-visual feedback. Full VoiceOver coverage now spans every screen: chat, Novel Writer, Soul, tools, MCP, memory, and settings. Section headers carry heading traits for rotor navigation. Verbose Announcements mode reads responses aloud on demand.
Free to Use. No Ads. Ever.
All local model features are completely free with no advertisements. Novel Writer, Soul, 40+ built-in tools, the document reader, voice transcription, image generation, and iCloud sync: all free when using local GGUF, MLX, or Apple Foundation models. No paywalls on core features. No surprise limits. Cloud API models and the MCP Marketplace require a subscription, but everything on-device is yours at no cost.
Privacy. Not as a Feature. As the Foundation.
Privacy AI runs entirely on-device. No backend servers. No accounts. No data collection. All local inference happens on your device - your conversations never leave. For remote API models, every request and response is fully visible in the built-in Protocol Inspector. Route through mitmproxy for independent verification.
No sign-ups. No tracking. You decide where execution happens, how data flows, and when tools are invoked.
Your models. Your device. Your data.
Terms of Use: https://privacyai.acmeup.com/docs/policy/tos.html
Privacy Policy: https://privacyai.acmeup.com/docs/policy/privacy.html
Comments
Perplexity and Copilot
Maybe you can get "me" to recommend that as well? thanks
Perplexity and Copilot
Maybe you can get "me" to recommend that as well? thanks
Check out the new vision assistant!
You might want to give the vision assistant and all the other new features a shot.
Where's the documentation?
In laymen's terms, hopefully. All the options were baffling to me. I didn't know which to choose. I also couldn't figure out where to enter my questions or how to send them. Maybe I was just having a bad day, but instructions in simple terms would be helpful to this 72 year old blind man.
I acknowledge the app and menus are relatively complicated
Double-tapping the Menu button should bring up a menu where you can access various functions of the app. Double-tapping on an expandable item expands it and you can then access the submenus. Double-tapping the expanded parent element again collapses it. Head to Menu > Models & Services > Local Models > GGUF Models and select Local AI to start chatting, or Apple Intelligence, assuming your device supports and you've enabled it. You can also download new and more capable models without selecting either of these options, and you may want to download MLX models instead of GGUF models. This app even supports MLX audio and TTS/STT, so it's too complicated to cover that simply. It's really an all-in-one AI workstation.
In-app purchase
I just have one question for now, what does the in-app purchase give us?
Letter letter letter
Thank you sincerely for offering a way for me to cut to the quick. I'll follow your advice on how to get started. Should I assume that if I want additional information regarding the app, I should just go ahead and do some google research on this feature and that? I apologize for my ignorance, but, unfortunately, much of what you wrote is still letter letter letter to me. Thank God you did give me a shortcut to at least get started! smile
Responding to Both Questions
@Brian you get a bunch of premium features including the ability to provide your API key and connect to remote servers so that you can chat with larger LLMs online. There's also this thing called MCP and I've not looked into it in detail as yet.
@Bruce Harrell I would be happy to help, but not each and every question on AI and LLMs, different model formats (GGUF, MLX etc.) and quants (UD_IQ3_XXS or MLX-4bit etc.), model architectures, the various trade-offs between different models like Qwen vs. Gemma, or hardware requirements can I answer here on this forum. You should refer to dedicated websites hosting useful articles and other resources on those. I've not learned everything at once either.