⬢github Go · 18 ★ +18 since we first saw it · pushed 1 h ago · MIT
Vibra-Ingenn/Janus
Janus is a API router for AI models written in Go and has a Vulkan Model runner
Janus is a single Go binary that runs local GGUF models using llama.cpp with Vulkan (AMD/Intel/NVIDIA) or CPU fallback, exposing an OpenAI-compatible API and a built-in web UI. It routes requests, runs built-in tools (file I/O, shell, OCR, PDFs), supports hot-swapping models, and optionally uses Ollama as a backend—no Python or Docker needed.
Why now: It was just shared on Hacker News as a 'Show HN' demo, likely appealing because it offers a dependency-free, cross-vendor-GPU alternative to Ollama for local LLM serving.
Who it is for: Developers and hobbyists who want to run local LLMs on any GPU (especially Windows machines) and plug them into OpenAI-compatible clients like Cursor or Cline.
Stars over our 11 snapshots: 0 to 18, since 2 h ago.
Where people talked about it
API: https://socialmediatrends-api.osmike.com/v1/repos/Vibra-Ingenn/Janus