AI Offline LLM Chat vs LLM API Request

Your ideas. Your choice of AI. Lapis is a native workspace for private AI chat, documents, images and voice on iPhone, iPad and Apple Silicon Mac. Download supported models to run locally, use Apple's on-device model when available, or explicitly choose Apple Intelligence Cloud for online requests. WRITE, LEARN AND EXPLORE Ask questions, refine writing, understand concepts and work through code. Keep work organized in projects with their own instructions. Duplicate a conversation or fork from a particular message without changing the original. Export chats, use temporary conversations, and protect saved chats with system authentication. DOCUMENTS AND PHOTOS With Pro, attach PDFs and text documents, or ask a compatible vision model about an image. Supported text formats include Markdown, CSV, JSON, XML and YAML. Scanned PDF pages use local text recognition, so even a text-only model can read a legible scan. Recognition may make mistakes; diagrams and visual layout are not included in text import. Documents must fit the model's context window. AN IMAGE STUDIO ON YOUR DEVICE Create locally with PrismML Bonsai Image 4B, choosing square, portrait or landscape. Or select Apple Image Playground on supported devices and refine your prompt in Apple's native interface. Keep creations in a local gallery, zoom, crop, copy prompts, regenerate, save and share. Select multiple images for batch actions. Bonsai requires a separate model download and enough available memory; Apple's system feature has its own availability and options. ENGLISH LIVE VOICE Talk naturally using the downloaded Liquid voice model, local speech recognition and spoken responses. Live Voice supports English only and requires compatible hardware and microphone/speech permissions. MODELS THAT FIT Browse supported local models, with download sizes, capabilities and conservative device compatibility guidance. Download recovery and verification help keep installed models usable. Apple Silicon Macs receive recommendations suited to their unified memory. On iPhone and iPad, New Chat and Live Voice widgets open the existing app flows directly. PRIVATE BY DESIGN, WITH A CLEAR ONLINE CHOICE Local model inference runs on your device. After setup, supported local features work offline. Conversations and generated images are stored locally. No account or third-party AI API key is needed for local AI. Apple Intelligence Cloud is optional and requires explicit permission, internet access, supported Apple software and service eligibility. Those requests are processed on Apple's Private Cloud Compute servers and are subject to Apple's limits. Apple Image Playground uses Apple's system framework; offline availability is not guaranteed. Model downloads and StoreKit purchases require internet access. LAPIS PRO Free includes selected models, up to 100 normal chats and 3 completed image generations in total. Optional monthly or annual Pro subscriptions unlock attachments, the complete supported model catalog, compatible model imports, Live Voice, projects, advanced controls, and chats and images without plan limits. Eligible new subscribers receive a 7-day free trial; the chosen plan renews at its displayed price unless cancelled. Restore or manage purchases through your Apple account. For accounts whose App Store country is Spain, Pro access is free without a subscription. Availability, context, speed and image resolution depend on your device, selected model and available memory. AI may make mistakes; verify important information. Privacy Policy: https://lapis.lvrpiz.com/privacy-policy Terms of Use: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/

LLM API Request lets you send requests to leading AI models straight from Apple Shortcuts, using your own API keys. No subscriptions. No accounts. SUPPORTED PROVIDERS • OpenAI • Anthropic • Google Gemini • xAI Grok • Perplexity KEY FEATURES • Bring your own API key — pay only for what you use (including generating images), at provider rates • Use local models like Gemma 4, and the Apple Intelligence model; Local models can even use web search! • Attach images, PDFs, and text files to your requests • Continue conversations across Shortcut runs with conversation history • Export any conversation as Markdown — great for apps like Obsidian • Track token usage across all models with a built-in dashboard • Export your token usage history • Use Projects to summarize documents and include them in the model's context. Donate the project info to Siri so that it can use that to answer your questions! • Test connections to ensure your key is working and browse available models per provider • Request structured JSON output for automation pipelines SECURITY FIRST Your API keys are stored exclusively in Apple Keychain, protected by Face ID or Touch ID. We run no servers and collect no data — everything stays on your device.