kamjin3086/chatless

What it solves

chatless is a local-first desktop AI chat client that allows users to interact with various AI models while keeping their data stored locally to protect privacy. It eliminates the need to rely on a single AI provider by supporting both cloud-based services and local model runtimes.

How it works

Built with Tauri 2.0 and Next.js, the application acts as a unified interface for multiple AI backends. It connects to cloud providers like OpenAI, Anthropic, and DeepSeek, or local runtimes such as Ollama and LM Studio. To enhance responses, it implements a local RAG (Retrieval-Augmented Generation) system that parses documents (PDF, Word, Markdown) and uses vector retrieval to provide context-aware answers.

Who it’s for

Users who want a professional, privacy-focused desktop interface for their AI workflows, especially those who combine cloud LLMs with local models and need to manage their own knowledge base via RAG.

Highlights

• Multi-Provider Support: Seamlessly switch between cloud APIs and local models.
• Local RAG Knowledge Base: Document parsing and vector search for more accurate, grounded responses.
• Vision Support: Ability to analyze images using Vision-capable models.
• MCP Protocol Integration: Extends the client's capabilities through third-party tool integration.
• Prompt Management: Tools to save, reuse, and sync prompts across devices via WebDAV.
• Local-First Storage: All user data is stored locally using SQLite for maximum privacy.