$ show lemonade-sdk/lemonade
Lemonade is a local AI inference server and desktop application that provides OpenAI/Anthropic/Ollama-compatible APIs with support for multiple backends (llama.cpp, vLLM, whisper.cpp, Kokoro, etc.). The C++ server core handles model lifecycle, routing policies, multi-modal inference, and telemetry, while the React/TypeScript frontend (Tauri desktop app and web build) offers a full GUI for chat, image/audio generation, transcription, and model management. The repository also includes a CLI client, tray application, extensive integration tests, and automated documentation generation. Architecture follows a clean backend-descriptor registry pattern with generated boilerplate, platform abstraction for process/archive/path operations, and a sophisticated routing policy engine for model selection.
The LLM identified 3 files worth highlighting for reuse:
Single-pass header-only GGUF metadata reader that extracts architecture, context length, KV cache parameters, and capability hints from GGUF model files. Implements low-level binary helpers for version 3+ format and is self-contained, making it reusable in any C++ project that needs to parse GGUF headers.
Expression language evaluator with tokenizer, parser, reference resolution, and condition evaluation. Designed for the job system but implemented as a standalone module with no external dependencies, suitable for embedding in any project needing a lightweight condition expression engine.
Standalone Python script that boots a live server, fetches backend descriptors via API, and injects generated content into Markdown/JS/JSON files using marker-delimited regions (BEGIN/END GENERATED). Supports a --check flag for CI drift detection and is directly reusable for any project that needs to keep documentation in sync with runtime metadata.