Privacy AI vs Cumbersome

Privacy AI is both a chatbot and an agent, and both run on your device. Ask it a question, or give it a goal and it will plan, use tools, and come back with a report. Local models work fully offline, and nothing leaves your phone unless you connect a cloud provider yourself. Agent - Give it a goal and it keeps working across as many turns as that takes, instead of answering once - It plans first, calls tools, keeps its own folder of files, and delivers a written report you can share and reopen later - Suited to work that has to be broken down first, such as deep research, a multi-leg trip, or comparing options - Teach it with Skills: folders of instructions and scripts it can call by name Chat - Set model tiers once (fast, capable, reasoning, local), then just type - Smart Chat routes each message by complexity and privacy, and keeps sensitive text local - Switch local and remote models mid-conversation, keeping context Run Models Locally or in the Cloud - On-device GGUF and MLX models, with vision and tool-calling - Connect to OpenAI, Claude, Gemini, DeepSeek, or any OpenAI-compatible server - Apple on-device Foundation Model (iOS 26+) System Keyboard - Bring Privacy AI into any app: dictate by voice, or Process to rewrite, translate, and reshape text in place - Build named pipelines (Polish then Translate) or use presets like Proofread and Email - Types in 11 languages with Pinyin, Japanese, and Korean input Mini Apps - Doc Scanner: turn a PDF, a photo, or a camera scan into clean Markdown, with tables and equations intact - Email Assistant: triage any IMAP inbox on device, with a dated report you can undo one tap at a time - Knowledge Wiki: a cross-linked knowledge base from your documents, with a graph and cited answers - LLM Council: query several models in parallel and synthesize one answer - Custom assistants with instructions, an avatar, and a knowledge base Tools - 60+ built-in tools: web search, news, calendar, code execution, and more - Full MCP support with a marketplace, Siri, and Shortcuts Writing and Voice - Long-form writing assistant that drafts outlines and chapters - Text to Audio: turn a document into a single audio file using on-device voices - Real-time voice chat and live transcription Files and Content - Convert PDFs, Office files, EPUB, video, and audio to Markdown, with OCR - Media Studio: generate images with your provider Accessibility - VoiceOver support, sound and haptic feedback, and Verbose Announcements - View Assistant: point the camera and hear what is in front of you Free to Use. No Ads. Local model features are free, including the AI keyboard, knowledge wiki, writing assistant, custom assistants, 60+ tools, document reader, and image generation. Cloud API models and the MCP marketplace need a subscription. Privacy by Design Privacy AI runs on-device. No backend servers. No accounts. No data collection. Every API request is visible in the Protocol Inspector. 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

IMPORTANT: Requires your own API keys from OpenAI, Anthropic, OpenRouter, Google AI Studio, Z.ai, or Vercel AI Gateway, or your own OpenAI-compatible endpoint. I built Cumbersome because consumer AI apps kept dumbing things down. Cumbersome is the manual transmission of AI apps: more complex, but with far greater control. WHY I BUILT THIS Consumer AI providers like ChatGPT and Claude route requests to cheaper models to keep their costs down. They charge monthly subscriptions whether you use them once or a thousand times. They hide the technical controls that let you tune how models respond. And they lock you into one provider. I wanted an AI client that connects directly to AI provider APIs. I wanted to pick exactly which model I use and know I'm getting it. I wanted to compare different models on the same prompt. I wanted to pay per use instead of subscriptions. So I built it. This is the app I use every day to interact with AI, so I pay attention and try to get it right. SUPPORTED PROVIDERS Bring your API keys, or add your own OpenAI-compatible endpoint, for any of the following: — ChatGPT / OpenAI — Claude / Anthropic — Gemini / Google AI Studio — OpenRouter multi-provider — Vercel AI Gateway multi-provider — Z.ai — OpenAI-compatible custom endpoints, including Ollama, Azure, LM Studio, Groq, Together, and self-hosted backends WHAT YOU GET — A free app. — A versatile API client. One app, every major provider and model. — Total control. Pick your model, tune the parameters, switch providers mid-conversation. — Compare models head to head. Stream multiple chats in parallel and let an AI judge pick the best. — Bring your own tools. Local browser use, reasoning, web search, image generation. — Rich, editable conversations. Markdown, LaTeX, code blocks, attachments, edit and export anything. — Private by design. iCloud sync across devices. Nothing saved to our servers. — Native and accessible on iPhone, iPad, and Mac. MAJOR FEATURES ADDED IN 2026 Highlights only. See What's New on the App Store for the latest. 2026-07-02: Chat appearance themes and reading fonts (Serif, Rounded, Mono). 2026-06-28: Folder-level system message presets. 2026-06-24: Prompt caching, with cache usage shown across providers. 2026-06-17: Reusable system message presets per conversation. 2026-06-16: Offload long system messages to iCloud Drive. 2026-06-15: Pause custom API parameters with a toggle. 2026-06-11: Compact a conversation to lower API costs. 2026-06-07: Run JavaScript and HTML directly from Cumbersome. 2026-06-04: Choose how OpenRouter parses PDFs and which search engine it uses. 2026-05-12: Configurable Quick Prompt shortcut and App Visibility on Mac. 2026-05-09: Conversation folders. Group, move, and organize chats. 2026-05-03: Collapsible sidebar sections, synced across devices. 2026-04-22: AI image generation with OpenAI and Gemini. 2026-04-18: Local browser tool use. AI can search and browse pages. 2026-04-11: Headless mode. The AI works the browser in the background. 2026-04-03: Searchable model picker for long model lists. 2026-03-15: Extended thinking and reasoning across providers. 2026-03-09: Auto-delete conversations after a set number of days. 2026-02-28: OpenAI-compatible custom providers (Ollama, Azure, LM Studio, Groq). 2026-02-25: Only Search When Requested. Keep web search off until you ask. 2026-02-24: Image paste, interactive attachment previews, and PDF thumbnails. 2026-02-21: Siri and Shortcuts integration. 2026-02-20: Vercel AI Gateway support, with model exclusion filters. 2026-02-06: Voice Control, Larger Text, and higher contrast. 2026-01-11: Major accessibility pass. VoiceOver, Dynamic Type, keyboard nav. 2026-01-06: OpenRouter multi-provider integration. THE DEAL The app is free. You pay providers directly for API access. I'm Peter, an indie developer in Reno, Nevada, out among the high desert tumbleweed. No VC money. No growth hacking. Just a tool I wanted to exist.