Help¶

Welcome to the Salut development-preview help. Salut has not been released. Pinned llama.cpp is the only inference backend, and GGUF is the only model format.

Topics¶

General Settings: Configure the port, display name, log level, API key, and other basic options.
Playground: Configure the chat playground hotkey, window behavior, and generation parameters.
Models: Understand exact GGUF identity, development deployment plans, and memory admission.
Peers & Trust: Learn about peer discovery, the pairing flow, and how to manage trusted peers.
Network: Understand participation modes, native multi-node execution, and cluster tokens.
API: Reference for the OpenAI-compatible API endpoints, streaming, and integration tips.
Security: How Salut handles identity, encryption, trust, and data privacy.
Troubleshooting: Solutions for common issues with discovery, connectivity, and performance.