QikLM

QikLM is web-based AI orchestration that runs on your local hardware. It connects directly to your llama.cpp or vLLM engines. There is nothing embedded. No Docker. No Python. No npm. Just one small binary. Features: Unified API: Run local models with cloud APIs like DeepSeek or GLM. Auto-Engine Boot: Auto-boot on prompt with no preloading. Chat Workspace: Full chat UI with web search, GPU monitoring, and LAN access. Integrations: Claude Code, Codex, and more. Privacy-first. Self-hosted. Qik.