Vibe Coding Discoverllmman
View on GitHubRun any agent on any model, models stored as OCI images
Rust CLI and server for running local or hosted models and connecting them to coding agents. Packages models as OCI artifacts, supports llama.cpp, vLLM, SGLang, and MLX-LM, and exposes compatible Ollama, OpenAI, and Anthropic APIs.
Use Cases
Serve local or hosted language models through a unified APIRun coding agents such as Claude Code, Codex, and OpenCode against local modelsStore models as OCI artifacts in standard registriesTransfer models directly from Hugging Face to OCI registriesPool inference capacity across multiple machinesGenerate images, video, and audio from diffusion modelsVerify model artifacts with cosign-compatible signatures
Built With
- Language
- Rust
- Frameworks
- llama.cpp · vLLM · SGLang · MLX-LM · Docker · Podman · Ollama API · OpenAI API · Anthropic API
Tags
local inference · LLM serving · OCI images · model registry · model transfer · agent integration · multi-provider · OpenAI-compatible API · Anthropic-compatible API · Ollama-compatible API · model signing · GPU inference · image generation · video generation