Lemonade — Local AI Server + CLI (Chat/Image/Speech)

Main

• One CLI, multiple modalities: README shows lemonade run for Gemma (chat), SDXL (image gen), Kokoro (speech), and Whisper (transcription).

• Built-in model library/manager: browse models and download via lemonade pull or the UI model manager.

• Developer-friendly switching: use lemonade backends to see available inference backends on your machine and pick what fits your hardware.

• App integrations: README links to a marketplace with multiple app connectors (including Claude Code docs link).

Source-backed notes

• README includes CLI commands: lemonade run, lemonade pull, lemonade list, lemonade backends, and lemonade launch claude.
• README states Lemonade supports multiple model formats (GGUF/FLM/ONNX) and multiple modalities (text, image, speech, transcription).
• README provides platform install entry points via the official install options page for Windows/Linux/macOS and Docker.

FAQ

• Do I have to use the UI?: No — README shows you can list/pull/run models entirely from the CLI.
• Can it handle non-chat workloads?: Yes. README includes image gen, speech gen, and transcription examples.
• How do I connect it to coding tools?: Use the integration docs/marketplace links; README also shows lemonade launch claude.