Dakodeon lives in your menu bar. Start a local OpenAI-compatible router, point your agent at it, and switch models by id from the client — no runtime to bundle, no config to babysit.
A built-in Settings window tracks every curated model. Watch downloads in real time and free up disk in a click — your agent switches models by id over the API, no picker to babysit.
Progress, sizes, and hashes resolve straight from Hugging Face file metadata — no guessing, no stale numbers.
OpenCode (or any client) sends a model id and the router loads that profile with its configured assets and flags. Dakodeon just shows what's live.
Delete weights you're done with. One confirm and the backing files leave the cache.
Serves llama-server at 127.0.0.1:8080/v1 with router presets for agents like OpenCode.
Uses the system llama.cpp and hf you already have. The app bundle stays tiny.
Each profile ships ready to run with curated weights, optional companion assets, and tuned flags. No knobs to get wrong.
Requires llama.cpp and hf on your PATH · macOS 14+