model-swapping

Run GGUF through llama.cpp and SafeTensors through vLLM behind one OpenAI-compatible endpoint. Your coding tools select a model; the switchboard manages the local runtime, process, and resident-model change.

Fast model suspend/restore for vLLM sleep mode — snapshot weights to disk and swap models on one GPU in seconds, not minutes.

SwapOS: frontier-class agentic AI pipelines on commodity hardware by swapping specialized small models (Context Capsule runtime)

Heuristic signals on whether an OpenAI-/Anthropic-compatible endpoint serves the model it claims — catch model-swapping, quantization & silent context truncation. Zero-dependency Python CLI. Signals, not proof.