Last released Jun 19, 2026
A single-port OpenAI- and Ollama-compatible reverse proxy that swaps the GPU between local LLM engines on demand.
Supported by