Versions:
Witsy 3.5.2, published by Nicolas Bonamy, is a desktop AI assistant positioned as a universal MCP (Model Context Protocol) client that lets users integrate virtually any large-language-model provider into a single, lightweight workspace. Operating on a BYOK (Bring Your Own Keys) basis, the application requires personal API credentials for cloud services such as OpenAI, Anthropic, Mistral AI, Google, xAI, Azure, OpenRouter, DeepSeek, Groq and Cerebras, while also offering zero-cost local inference through Ollama. Beyond conventional chat completion, the program supports vision models for image description, text-to-image and text-to-video generation via OpenAI, Google, Replicate, fal.ai and HuggingFace, plus image-to-image editing and image-to-video conversion. An expandable plugin architecture allows Python code execution, web searching and other augmentations, and full compatibility with Anthropic MCP servers enables advanced tool chaining. Productivity features include a scratchpad for iterative prompting, “Prompt anywhere” hotkeys that inject AI output directly into any active window, AI commands that operate on selected text, expert prompt libraries for domain-specific bots, long-term memory for contextual continuity, and RAG-based conversation with local documents. Speech capabilities cover dictation, transcription and read-aloud of either assistant messages or arbitrary on-screen text through OpenAI or ElevenLabs voices. With thirty-five released versions to date, Witsy continues to evolve as a vendor-neutral hub for generative tasks ranging from creative writing and software development to multimedia production and knowledge retrieval. The software is available for free on get.nero.com, with downloads provided via trusted Windows package sources (e.g. winget), always delivering the latest version, and supporting batch installation of multiple applications.
Tags: