Local GGUF runtime
Run compatible GGUF models through bundled llama.cpp runtimes for Mac and Windows. The installer stays model-free; users import their own models.
Private desktop AI model runner
Realigns AI Studio 1.0 is a private local AI workspace by Realigns Inc. It helps users run compatible GGUF models locally, manage prompts, process documents, use controlled web search, and connect local tools through a private API gateway.
Professional capabilities
Realigns AI Studio 1.0 combines local model execution, structured prompts, document assistance, optional web research, and local API access without forcing users into a cloud-first workflow.
Run compatible GGUF models through bundled llama.cpp runtimes for Mac and Windows. The installer stays model-free; users import their own models.
A focused chat interface with conversation flow, copy actions, model status, web result cards, and professional response formatting.
Save and edit system prompts for each model so business, general, research, and assistant modes can follow a consistent style.
Upload supported local documents such as TXT, Markdown, PDF, and DOCX for private document-grounded assistance.
Enable web search when needed for source cards, image cards, and research support. Keep it disabled for fully offline/private use.
Generate local API keys and connect scripts, internal tools, websites, and office workflows to the local OpenAI-compatible endpoint.
Application sections
Replace these Unsplash placeholders with your real screenshots from Home, Models, Chat, Prompts, Developer API, Settings, and Policies pages.

A clear product landing area inside the app with device hardware detection, privacy summary, and quick access to models and chat.

Users can add custom GGUF files, choose model mode, set context size, temperature, max tokens, and start or stop the runtime.

The chat workspace supports local conversation, selected documents, source cards, and controlled web research when enabled.

Local API keys, offline/online controls, web search settings, privacy notice, and generation parameters are managed from one place.
Privacy-first architecture
Realigns AI Studio is designed so local chats, imported GGUF models, prompts, documents, and AI outputs remain on the user's computer during local/offline use. Online features are optional and clearly separated.
Runtime package
Version 1.0 includes local runtime support for both platforms. Mac uses bundled Darwin ARM64 runtime libraries, while Windows includes the required llama-server.exe and runtime DLLs.
| Platform | Included runtime |
|---|---|
| Mac ARM64 | dmg. |
| Windows x64 | zip. |
How users work
Use the Mac DMG or Windows setup file. The app installs without bundling large GGUF models.
Users review the local privacy promise, online feature notice, commercial license, and security responsibilities.
Users browse for their local model file and define display name, family, mode, context size, and prompt template.
The backend and llama runtime select local ports automatically and show model-ready status when loading is complete.
Users can chat locally, tune prompts, attach documents, enable controlled web search, or connect tools through the local API gateway.
Commercial release notice
Realigns AI Studio 1.0 is distributed with commercial release documentation covering permitted use, restrictions, local privacy promise, online feature notice, model compatibility, support, and user responsibilities. Users must ensure that imported models are legally obtained and suitable for their hardware.
Apple Silicon DMG package for private commercial testing and installation.
Windows setup EXE and ZIP package with bundled Windows llama runtime.
Commercial license, privacy/security notice, model compatibility notice, release README, and checksum file.