TangiLM vs WristMark

TangiLM — Run AI Locally on Your iPhone TangiLM is a mobile language model app that lets you run powerful chat and reasoning locally on your iPhone. No account required, no cloud subscription, and no need to send your prompts to remote servers. Download your own GGUF models easily, manage them easily, and chat with full control over what runs on-device. Key Features Offline-first private chat Run your language model locally for everyday conversations, brainstorming, Q&A, and writing. Keep your data on your device to support a more private workflow. Download and manage local models Browse, download, and load your favorite GGUF models from Hugging Face. The app scans your model directory, tracks download progress, and lets you switch models anytime. Optional web search for up-to-date answers When enabled, TangiLM can perform web searches and pull relevant page content to help answer time-sensitive questions. This hybrid approach keeps casual chat fast while still supporting more current information when you choose. Built-in search workflow (analyzing → searching → evaluating → fetching) For improved relevance, the app uses an advanced internal agent loop to refine search queries, decide whether to fetch a page, and combine results into the final context used for generation. Thinking and searching controls If your selected model supports advanced reasoning, you can enable a “thinking” mode for better step-by-step answers. You can also toggle the web search feature per conversation. Fine-tuned performance and stability TangiLM also lets you fine-tune inference parameters for better speed and stability. For newer model releases, the app provides strong compatibility and optimized inference behavior, which helps reduce the likelihood of overheating and crashes commonly seen in other iOS AI tools. Session and history support Your chat sessions are saved automatically so you can continue where you left off. Review messages, edit user content, and keep your workflow organized. Multilingual UI and prompts The interface is localized across multiple languages, and the app supports multilingual prompts for smoother international use. Why TangiLM Whether you’re building a private assistant, experimenting with different models, or asking questions that benefit from fresh web information, TangiLM gives you a practical, flexible toolkit: local speed and privacy by default, with optional online search when you need it.

WristMark is the private meeting companion built for iPhone and a paired wrist app. Record from your wrist, mark critical moments with a simple gesture, transcribe offline, and summarize with optional on-device Local AI or the AI endpoint you configure. WRIST-FIRST RECORDING & TIMESTAMPS Meetings are not the place to fumble with your phone. Start recording from the wrist app with one tap. When something important is said, use a quick tap gesture to plant an instant timestamp — hands-free, screen-free, and silent. During playback on your iPhone, jump to any marked moment in a single tap. No scrubbing. No guessing. ON-DEVICE TRANSCRIPTION — NO INTERNET REQUIRED Transcription runs on device using system speech technology. No audio is ever sent to a server by WristMark. Works fully offline. When the operating system improves on-device speech models, simply re-transcribe existing recordings to benefit from the latest accuracy gains. VOICE MEMO IMPORT No wrist recording available? No problem. Import recordings directly from Voice Memos using the system Share Sheet. Every imported recording gets full transcription, timeline view, and archiving — everything WristMark offers. MULTI-SEGMENT MERGE & ARCHIVE Interrupted by a call? Took a break mid-meeting? Recorded the afternoon session separately? WristMark lets you merge multiple audio segments into a single meeting archive. Organize by project or date. Clean, searchable, and friction-free. FASTER, MORE RESPONSIVE HOME Get back to your meetings sooner. WristMark shows the first part of your meeting history without waiting for the entire archive to load, keeps meeting actions responsive while the rest is recovered in the background, and makes the first recording easier to start when your home is empty. New meetings, summaries, and merged archives stay reflected in the home screen as they become available. LOCAL AI SUMMARIZATION Use the optional Local AI model for supported on-device summaries after downloading it in Settings, or connect any OpenAI-compatible API endpoint — your company's private LLM, a self-hosted model, or a third-party key you control. Local AI runs on your device. API requests are initiated by you and sent directly to your chosen endpoint. For summaries, transcript polish, or classification through a configured endpoint, WristMark sends transcript text, markers, meeting duration, and recording type. Audio files are not sent to the AI endpoint by WristMark, and meeting data never passes through WristMark's servers. WHO IS WRISTMARK FOR? - Professionals in high-frequency meeting environments - Students attending lectures, training sessions, or seminars - Enterprise users with strict data confidentiality requirements - Users who want fast wrist-first capture PRIVACY & SECURITY - Transcription runs on device - Optional Local AI summaries run on device after model download - Audio files are never sent to WristMark servers - Third-party AI summaries use user-initiated requests to your own API endpoint - No analytics, no user tracking, no account required LEGAL Terms of Use (EULA): https://wristmark-app.github.io/legal/terms.html Privacy Policy: https://wristmark-app.github.io/legal/privacy.html