The all-in-one AI workspace that runs on your machine. Chat, voice, image generation, agents, and more — fully offline.
View all releases on GitHub · Runs 100% offline · No subscriptions
A glimpse of what's waiting inside XandSuite V1
The model thinks through its plan step by step, then uses built-in tools — like PDF creation — directly from the chat window. No switching apps, no copy-pasting.
Describe what you want changed and the model re-runs the ComfyUI workflow with your edits. Updated images appear in the same conversation, ready to iterate on.
Describe your scene in plain language. XandSuite routes your prompt through ComfyUI and delivers the result directly in the chat window — entirely offline.
Turn a still image into an animated clip with a single message. The video is generated locally via ComfyUI and plays back right inside your conversation.
Build distinct AI personalities — each with their own avatar, system prompt, and model assignment. Switch between them per conversation to match any task or tone.
24 tools available across installed packages — image generation, video creation, media browsing, chart rendering, and PDF processing — all callable directly from the chat input.
Extend XandSuite with one-click installs. Official packages cover media servers, creative generation, data presentation, and document tools — each configurable and removable at any time.
Browse thousands of GGUF models from HuggingFace, download them with a single click, and run them locally. No API keys. No data leaving your machine.
Fine-tune every aspect of local inference — GPU layers, context size, thread count, and auto-detected CUDA binaries — all from a single, clean settings panel.
A growing library of reusable prompts across Creative, Code, Media, and Productivity. Pick one, fill in the variables, and send — or build your own and save them for later.
Tap the orb and speak. XandSuite listens with Whisper, processes with your local model, and responds aloud with KokoroTTS. A complete voice AI loop — no internet required.
22 capability areas in a single desktop application. Everything runs on your hardware — private, fast, and always available.
Install XandSuite, pick a model, and you're ready to go — no accounts, no API keys, no internet required.
Introducing XandLLM - a production-grade LLM inference engine written in Rust. OpenAI-compatible API, GPU acceleration, and built-in knowledge distillation for creating smaller, faster models.
How YouTube's deletion of Rewind videos highlights the importance of self-hosting media. A guide to preserving digital content before it disappears forever.
Complete guide to setting up Meetily for offline meeting transcription and AI summarization using Ollama with Gemma 3 4B. No cloud required, 100% local processing.