llm-local

llm-local is a lightweight abstraction layer for local Large Language Models (LLMs). It provides a unified API and a command-line interface (CLI) to interact with various local model providers, simplifying the process of switching between different backends. Key capabilities include: - Unified API: A consistent interface for generating text and streaming responses across different providers. - Provider Discovery: Automatic discovery and listing of available local providers and their models. - CLI Tool: A ready-to-use `llm` command that supports multi-turn conversation history and a toggleable 'think' mode. - Supported Providers: Out-of-the-box support for Ollama (via `/api/chat`) and LM Studio (via `/api/v1/chat`), with the ability to register custom provider instances. - Flexible Integration: Can be installed as a global CLI tool or as a dependency in TypeScript/JavaScript projects.