# Lemonade Server

Lemonade Server is a free, private local AI runtime that serves optimized models for chat, coding, speech, and image generation through OpenAI, Anthropic, and Ollama-compatible APIs, running entirely on your own GPUs. It is open source and self-hosted, installable on Proxmox.

Canonical page: https://ourlocalhost.com/app/4537/lemonade-server

## Description

Lemonade Server is a free, private local AI runtime that serves optimized models for chat, coding, speech, and image generation through OpenAI, Anthropic, and Ollama-compatible APIs, running entirely on your own GPUs and NPUs.

## Platform availability

Available on 3 of 10 platforms: Proxmox, Mac, Windows.

## Desktop and other devices

Mac, Windows.

## Health score

82/100 (Excellent), based on 4 of 5 factors.
- Maintained: 100 (last commit 5 days ago)
- Popular: 76 (6,241 GitHub stars)
- Easy to install: 45 (3 install platforms, install notes)
- Security check: 88 (no serious findings, 2 minor, CI, tests)

## Links

- GitHub: https://github.com/lemonade-sdk/lemonade
- Official website: https://lemonade-server.ai/
- desktop: https://github.com/lemonade-sdk/lemonade/releases/latest
- Proxmox: https://community-scripts.github.io/ProxmoxVE/scripts?id=lemonade-server
- Documentation: https://lemonade-server.ai/docs/api/

## Common questions

### How do I install Lemonade Server on Ubuntu?

Install it with the Snap package using sudo snap install lemonade-server. If this step is skipped, no Lemonade server is running, so its local API, models, and UI are unavailable.

### Why does model loading fail even after the server is installed?

Check the model's recipe and install a matching backend first, such as llamacpp:vulkan for compatible graphics hardware. If the required backend is missing, the model can fail to load even though Lemonade Server itself installed correctly.

### What causes an 'Origin not allowed' error in Lemonade Server?

It is related to CORS or API key configuration. Not setting LEMONADE_API_KEY can expose the server to unauthenticated access and cross-origin attacks, so wildcard CORS should only be used in secure development.

### Can I run Lemonade Server on one machine and use it from another?

Yes, Lemonade supports running the server on one machine while using the app from another machine on the same network. Check the remote server connection guide for instructions and security considerations.

### Does Lemonade use an AMD NPU automatically?

If the server detects your NPU, it can work automatically. On Linux you may still need to install the required package manually, while Windows can be easier to set up.

### Is there a mobile app for Lemonade?

Yes, Lemonade offers iOS and Android apps. The mobile app source is also available.

## Alternatives

- AI Playground: This application provides a full suite of generative AI features for chat, code assistance, document search, image analysis, image and vide…
- AnythingLLM: AnythingLLM is a local, all-in-one AI app for self-hosters that works with any large language model, providing document chat, retrieval-aug…
- Ollama: Ollama is a local runtime for large language models (LLMs) that provides an application programming interface (API) and command-line interf…
- llama.cpp: Llama.cpp is a pure C/C++ inference engine for large language models, including Meta's LLaMA.
- oMLX: LLM inference server with continuous batching &amp; SSD caching for Apple Silicon — managed from the macOS menu bar.

---
Source: Our Local Host (https://ourlocalhost.com/app/4537/lemonade-server), the self-hosted and open-source app directory.
