Local AI Chat vs Raven Documents

SmartFruit provides a chat interface to test Apple Intelligence's local language model on the device. It works offline, thereby protecting your privacy. The on-device model is supports a wide range of text generation tasks, like summarization, entity extraction, text understanding, refinement, dialog for games, generating creative content, and more. World knowledge to a lesser extent. The quality of responses in this area is naturally not on par with large cloud-based models, but that’s beyond my control. Especially developers can try out all generation options of Apple's Foundation Models framework and get a corresponding code example.. Furthermore, chats can be archived for later review after a one-time in-app purchase. Important: This app requires iOS 26 and Apple Intelligence. Please ensure your device is compatible before purchasing.

Raven is your private open source AI assistant for iOS. It helps you work with documents, images, audio, and video files directly on your device. Nothing is sent to external servers. All processing is done locally using Apple’s Foundation Models. Raven is ideal for working with small to medium-sized files such as articles, notes, reports, transcripts, and PDFs. - Best suited for documents of moderate length - Handles long text comfortably while still providing clear summaries and answers - Designed to be fast and efficient without slowing down your device What you can do with Raven: - Extract text from images - Read and analyze PDF files - Transcribe audio files - Extract text from video content - Ask questions about your files - Summarize and analyze content - Organize chats and manage files Privacy first Your data never leaves your iPhone. Raven runs fully on your device with no cloud services and no internet required for AI features. - Designed for productivity - Create and manage multiple projects - Upload and organize your files - Take notes automatically saved - Switch between chats with ease Built for iOS Raven is made for iOS and takes full advantage of Apple’s frameworks including Foundation Models, Vision, PDFKit, Speech and AVFoundation. The full source code is available on GitHub.