Lemonade

https://lemonade-server.ai/docs
Admin
Lemonade is a local AI platform that makes running and embedding AI models simple and enjoyable,...
Tokens:97,585
Snippets:1,260
Trust Score:8.3
Update:2 months ago
Show doc for...
Tokens:
Raw
### Run Setup Script

Source: https://lemonade-server.ai/docs/dev/getting-started

Initializes the build environment and installs dependencies.

```bash
./setup.sh
```

```powershell
./setup.ps1
```

--------------------------------

### Run Lemonade Tools Example

Source: https://lemonade-server.ai/docs/dev/lemonade-omni

Commands to install dependencies and execute the end-to-end tool loop example.

```bash
pip install openai
python examples/lemonade_tools.py "Generate an image of a sunset"
python examples/lemonade_tools.py "Say hello world out loud"
```

--------------------------------

### Local Documentation Preview

Source: https://lemonade-server.ai/docs/dev/documentation

Commands to install documentation dependencies and start the local development server.

```bash
pip install -r docs/assets/docs_requirements.txt
zensical serve
```

--------------------------------

### Install Backend (Packaging Time)

Source: https://lemonade-server.ai/docs/embeddable/backends

Installs a specified backend for a given recipe and device. This is useful for bundling backends into your application's installer.

```APIDOC
## lemonade backends install BACKEND:DEVICE

### Description
Installs a specified backend for a given recipe and device. This command is used to bundle backends into your application's installer.

### Usage
```bash
./lemonade backends install llamacpp:vulkan --force
```

### Parameters
* `BACKEND:DEVICE`: The backend to install (e.g., `llamacpp:vulkan`).
* `--force`: Ignores compatibility checks, useful for packaging on a VM.
```

--------------------------------

### Install Lemonade Server via Snap

Source: https://lemonade-server.ai/docs/guide/install/ubuntu

Install the Lemonade server using the snap package manager. This is a convenient way to get the latest version in a sandboxed environment.

```bash
sudo snap install lemonade-server

```

--------------------------------

### Install wut-cli

Source: https://lemonade-server.ai/docs/integrations/wut

Install the wut CLI tool using pipx.

```bash
pipx install wut-cli
```

--------------------------------

### Install Backend (Runtime)

Source: https://lemonade-server.ai/docs/embeddable/backends

Installs a specified backend for a given recipe. This is used when the backend choice depends on the user's hardware.

```APIDOC
## POST /v1/install

### Description
Installs a specified backend for a given recipe. This is used when the backend choice depends on the user's hardware.

### Method
POST

### Endpoint
/v1/install

### Parameters
#### Request Body
- **recipe** (string) - Required - The recipe for the backend (e.g., "llamacpp").
- **backend** (string) - Required - The specific backend to install (e.g., "rocm", "vulkan").
- **stream** (boolean) - Optional - Whether to stream the installation process.

### Request Example
```json
{
  "recipe": "llamacpp",
  "backend": "rocm",
  "stream": false
}
```

### Response
#### Success Response (200)
(No specific fields mentioned in the source, typically indicates successful initiation of installation)

#### Response Example
(No example provided in source)
```

--------------------------------

### Bot Command Examples

Source: https://lemonade-server.ai/docs/integrations/mindcraft

Examples of how to format bot responses using specific commands like !goToPlayer, !startConversation, and !endConversation.

```text
On my way! !goToPlayer("zZZn98", 3)
System output: Arrived at player.
Your output:
Here!
User input: zZZn98: no come right where I am
Your output:
Okay, I'll come right to you. !goToPlayer("zZZn98", 0)
```

```text
!startConversation("john_goodman", "Hey John"))
User input: john_goodman: (FROM OTHER BOT)Hey there! What's up?
Your output:
Hey John, not much. Just saying hi.
User input: john_goodman: (FROM OTHER BOT)Bye!
Your output:
Bye! !endConversation('john_goodman')
```

```text
On my way! !goToPlayer("Transpier", 3)
```

--------------------------------

### Install .deb package

Source: https://lemonade-server.ai/docs/dev/getting-started

Install the Debian package using apt.

```bash
# Replace <VERSION> with the actual version (e.g., 9.0.0)
sudo apt install ./lemonade-server_<VERSION>_amd64.deb
```

--------------------------------

### Install Moonshine Backend

Source: https://lemonade-server.ai/docs/guide/configuration/moonshine

Install the Moonshine CPU backend using the CLI or the HTTP API.

```bash
lemonade backends install moonshine:cpu
```

```bash
curl -X POST http://localhost:13305/api/v1/install \
  -H 'Content-Type: application/json' \
  -d '{"recipe": "moonshine", "backend": "cpu"}'
```

--------------------------------

### Install lemonade-desktop

Source: https://lemonade-server.ai/docs/guide/install/arch

Installs the optional desktop frontend package for the lemonade server.

```bash
pacman -S lemonade-desktop
```

--------------------------------

### Common Lemonade CLI Commands

Source: https://lemonade-server.ai/docs/guide/cli

Examples of standard operations including model management, backend installation, and data export.

```bash
# List all available models
lemonade list

# Pull a custom model with specific checkpoint
lemonade pull user.MyModel --checkpoint main org/model:Q4_K_M --recipe llamacpp

# Load a model with custom recipe options
lemonade load Qwen3-0.6B-GGUF --ctx-size 8192

# Install a backend for a recipe
lemonade backends install llamacpp:vulkan

# Register a cloud OpenAI-compatible provider (see `lemonade cloud --help`)
lemonade cloud install fireworks --base-url https://api.fireworks.ai/inference/v1

# Export model info to JSON file
lemonade export Qwen3-0.6B-GGUF --output model-info.json
```

--------------------------------

### Install Lemon Zest on Linux via Snap

Source: https://lemonade-server.ai/docs/integrations/lemon-zest

Install the application using the Snap package manager or build the snap package locally.

```bash
sudo snap install lemon-zest
```

```bash
sudo snap install snapcraft --classic
snapcraft
sudo snap install lemon-zest_*.snap --dangerous
```

--------------------------------

### Build and Run Commands

Source: https://lemonade-server.ai/docs/guide/install/docker

Commands to build the image and start the containerized service.

```bash
docker-compose build
```

```bash
docker-compose up -d
```

--------------------------------

### Run Lemon Zest from Source

Source: https://lemonade-server.ai/docs/integrations/lemon-zest

Clone the repository and start the application using npm.

```bash
git clone https://github.com/phqen1x/lemon-zest.git
cd lemon-zest
npm install
npm start
```

--------------------------------

### Enable and Start Lemonade Service

Source: https://lemonade-server.ai/docs/guide/install/fedora

Configures the lemond service to start automatically on boot and starts it immediately.

```bash
sudo systemctl enable --now lemond
```

--------------------------------

### Install Vulkan backend on Windows

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the Vulkan backend for llama.cpp on Windows as a fallback if ROCm is not supported. The `stream: false` option indicates that the installation should not be streamed.

```bash
curl -X POST http://localhost:8000/v1/install ^
  -H "Content-Type: application/json" ^
  -d "{\"recipe\": \"llamacpp\", \"backend\": \"vulkan\", \"stream\": false}"
```

--------------------------------

### Start Lemonade Server

Source: https://lemonade-server.ai/docs/integrations/lemon-zest

Initialize the server with the required Flux-2-Klein-4B model before launching the application.

```bash
lemonade run Flux-2-Klein-4B
```

--------------------------------

### Install llama.cpp backend on Windows

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the llama.cpp backend with Vulkan support on Windows. Ensure `lemond` is running and downloading to the specified directory.

```batch
REM Start lemond to download backends to ./bin/
lemond.exe ./

REM Download llama.cpp with the Vulkan backend to ./bin/llamacpp/vulkan
lemonade.exe backends install llamacpp:vulkan
```

--------------------------------

### Install Cloud Provider

Source: https://lemonade-server.ai/docs/guide/configuration/cloud

Register a new cloud provider with the Lemonade server using the install command.

```bash
lemonade cloud install fireworks --base-url https://api.fireworks.ai/inference/v1
```

--------------------------------

### Install Vulkan backend on Linux

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the Vulkan backend for llama.cpp on Linux as a fallback if ROCm is not supported. The `stream: false` option indicates that the installation should not be streamed.

```bash
curl -X POST http://localhost:8000/v1/install \
  -H "Content-Type: application/json" \
  -d '{
    "recipe": "llamacpp",
    "backend": "vulkan",
    "stream": false
  }'
```

--------------------------------

### Install Dependencies

Source: https://lemonade-server.ai/docs/integrations/langchain

Install the necessary Python packages for LangChain and document processing.

```bash
pip install langchain langchain-openai langchain-community langchain-chroma langchain-text-splitters pypdf
```

--------------------------------

### Run Python CLI Example

Source: https://lemonade-server.ai/docs/api/openai

Execute the provided Python script to stream audio from a microphone.

```bash
# Stream from microphone
python examples/realtime_transcription.py --model Whisper-Tiny
```

--------------------------------

### Reference Path Examples

Source: https://lemonade-server.ai/docs/dev/job-expression-language

Examples of valid reference paths using the ${path} syntax to access context data.

```text
${model}
${run_v.timings.predicted_per_second}
${inputs.ctx}
${list.0}
```

--------------------------------

### Install a cloud provider

Source: https://lemonade-server.ai/docs/guide/cli

Registers a new provider with a base URL and optional API key.

```bash
lemonade cloud install PROVIDER --base-url URL [--api-key KEY]
```

```bash
lemonade cloud install fireworks --base-url https://api.fireworks.ai/inference/v1
```

--------------------------------

### Install Lemonade Package

Source: https://lemonade-server.ai/docs/guide/install/debian

Installs the downloaded .deb package using apt.

```bash
sudo apt install ./lemonade-server_*-debian13_amd64.deb
```

--------------------------------

### POST /v1/install

Source: https://lemonade-server.ai/docs/embeddable/runtime

Installs or updates a backend.

```APIDOC
## POST /v1/install

### Description
Installs or updates a backend.

### Method
POST

### Endpoint
/v1/install
```

--------------------------------

### Launch with Interactive Menu

Source: https://lemonade-server.ai/docs/integrations/claude-code

Starts the interactive flow for users to discover, import, or select models.

```bash
lemonade launch claude
```

--------------------------------

### Install LangChain dependencies

Source: https://lemonade-server.ai/docs/integrations/langchain

Install the necessary LangChain and OpenAI compatibility packages.

```bash
pip install langchain langchain-openai
```

--------------------------------

### Install pipx

Source: https://lemonade-server.ai/docs/integrations/wut

Install pipx to manage Python CLI tools on Linux.

```bash
sudo apt update
sudo apt install pipx
pipx ensurepath
```

--------------------------------

### Install lemonade-server

Source: https://lemonade-server.ai/docs/guide/install/arch

Installs the core lemonade-server package via the Arch Linux package manager.

```bash
pacman -S lemonade-server
```

--------------------------------

### Install Lemonade on Linux

Source: https://lemonade-server.ai/docs/dev/documentation

Installs the Lemonade package on Linux systems using apt.

```bash
sudo apt install lemonade
```

--------------------------------

### Install Pi Coding Agent

Source: https://lemonade-server.ai/docs/integrations/pi

Global installation command for the Pi coding agent.

```bash
npm install -g @earendil-works/pi-coding-agent
```

--------------------------------

### Advanced Model Registration Examples

Source: https://lemonade-server.ai/docs/guide/configuration/custom-models

Examples for registering models with multiple checkpoints, vision components, or custom labels.

```bash
# Register and pull a custom GGUF model with a main checkpoint
lemonade pull user.Phi-4-Mini-GGUF \
    --checkpoint main unsloth/Phi-4-mini-instruct-GGUF:Q4_K_M \
    --recipe llamacpp

# Register and pull a vision model with main + mmproj
lemonade pull user.Gemma-3-4b \
    --checkpoint main ggml-org/gemma-3-4b-it-GGUF:Q4_K_M \
    --checkpoint mmproj ggml-org/gemma-3-4b-it-GGUF:mmproj-model-f16.gguf \
    --recipe llamacpp

# Register a model with multiple labels
lemonade pull user.MyCodingModel \
    --checkpoint main org/model:Q4_0 \
    --recipe llamacpp \
    --label coding \
    --label tool-calling
```

--------------------------------

### Manage backends via CLI

Source: https://lemonade-server.ai/docs/guide/cli

Commands for listing, installing, and uninstalling backend recipes.

```bash
lemonade backends
lemonade backends --all
lemonade backends install SPEC [--force]
lemonade backends uninstall SPEC
```

```bash
# List supported recipes and backends
lemonade backends

# List all available recipes and backends
lemonade backends --all

# Install Vulkan backend for llamacpp
lemonade backends install llamacpp:vulkan

# Uninstall CPU backend for llamacpp
lemonade backends uninstall llamacpp:cpu

# Install FLM backend
lemonade backends install flm:npu

# Install an otherwise filtered backend
lemonade backends install llamacpp:rocm --force
```

--------------------------------

### Install Open WebUI

Source: https://lemonade-server.ai/docs/integrations/open-webui

Install the Open WebUI package within a dedicated Python environment.

```bash
pip install open-webui
```

--------------------------------

### Cloud Model Naming Examples

Source: https://lemonade-server.ai/docs/guide/configuration/cloud

Examples of dot-namespaced model identifiers for cloud providers.

```text
fireworks.accounts/fireworks/models/kimi-k2p5
fireworks.qwen3-235b-a22b
```

--------------------------------

### Install llama.cpp backend on Linux

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the llama.cpp backend with Vulkan support on Linux. The `--force` option bypasses system compatibility checks.

```bash
# Start lemond to download backends to ./bin/
./lemond ./

# Download llama.cpp with the Vulkan backend to ./bin/llamacpp/vulkan
./lemonade backends install llamacpp:vulkan --force
```

--------------------------------

### Start Lemonade Server

Source: https://lemonade-server.ai/docs/dev/getting-started

Commands to launch the server with default or custom port configurations.

```bash
./lemond
```

```bash
./lemond --port 8080
```

--------------------------------

### Verify installation

Source: https://lemonade-server.ai/docs/dev/getting-started

Check that the executables are available in the system PATH.

```bash
lemonade --help
lemond --help
```

--------------------------------

### Pull Model Examples by Registry

Source: https://lemonade-server.ai/docs/guide/configuration/custom-models

Examples of pulling specific model variants from Hugging Face and ModelScope.

```bash
# Hugging Face (shipped default source)
lemonade pull unsloth/Qwen3-8B-GGUF:Q4_K_M

# ModelScope
lemonade pull --source modelscope Qwen/Qwen3-8B-GGUF:Q4_K_M
```

--------------------------------

### Launch the chat REPL

Source: https://lemonade-server.ai/docs/guide/cli-chat

Basic command to start the interactive chat session.

```bash
lemonade chat
```

--------------------------------

### Start tmux Session

Source: https://lemonade-server.ai/docs/integrations/wut

Initialize a tmux session required for wut functionality.

```bash
tmux
```

--------------------------------

### Install Provider with API Key in One Step

Source: https://lemonade-server.ai/docs/guide/configuration/cloud

Register a provider and provide the API key directly in a single command.

```bash
lemonade cloud install fireworks \
  --base-url https://api.fireworks.ai/inference/v1 \
  --api-key fw-XXXXX
```

--------------------------------

### List cloud providers

Source: https://lemonade-server.ai/docs/guide/cli

Displays all installed providers, their status, and discovered models.

```bash
lemonade cloud list
```

--------------------------------

### Build and Notarize macOS Package

Source: https://lemonade-server.ai/docs/dev/getting-started

Commands to generate the signed installer and submit it to Apple for notarization.

```bash
cd src/cpp/build
cmake --build . --config Release --target package-macos
cmake --build . --config Release --target notarize_package
```

--------------------------------

### Get System Info

Source: https://lemonade-server.ai/docs/embeddable/backends

Retrieves information about the system's supported backends and their states. This is crucial for determining which backends can be installed.

```APIDOC
## GET /v1/system-info

### Description
Retrieves information about the system's supported backends and their states. This is crucial for determining which backends can be installed.

### Method
GET

### Endpoint
/v1/system-info

### Response
#### Success Response (200)
- **recipes** (object) - Information about available recipes and their backends.
  - **llamacpp** (object)
    - **backends** (object)
      - **rocm** (object)
        - **devices** (array) - List of supported ROCm devices.
        - **state** (string) - Installation state of the ROCm backend ('installable', 'installed', etc.).
      - **vulkan** (object) - Information about the Vulkan backend.

### Response Example
```json
{
  "recipes": {
    "llamacpp": {
      "backends": {
        "rocm": {
          "devices": ["amd_igpu"],
          "state": "installable"
        }
      }
    }
  }
}
```
```

--------------------------------

### Configure Custom Vulkan Binary Path (Linux)

Source: https://lemonade-server.ai/docs/embeddable/backends

Use these commands in bash to start `lemond` and then set the path to your custom Vulkan `llama-server` binary.

```bash
# Start lemond to update configuration
./lemond ./

# Set the llama-server vulkan binary path
./lemonade config set llamacpp.vulkan_bin /path/to/bins

```

--------------------------------

### Install ROCm backend on Linux

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the ROCm backend for llama.cpp on Linux if ROCm is supported by your system. The `stream: false` option indicates that the installation should not be streamed.

```bash
curl -X POST http://localhost:8000/v1/install \
  -H "Content-Type: application/json" \
  -d '{
    "recipe": "llamacpp",
    "backend": "rocm",
    "stream": false
  }'
```

--------------------------------

### Configure Custom Vulkan Binary Path (Windows)

Source: https://lemonade-server.ai/docs/embeddable/backends

Use these commands in Windows cmd.exe to start `lemond` and then set the path to your custom Vulkan `llama-server` binary.

```batch
REM Start lemond to update configuration
lemond.exe ./

REM Set the llama-server vulkan binary path
lemonade.exe config set llamacpp.vulkan_bin C:\path\to\bins

```

--------------------------------

### Install ROCm backend on Windows

Source: https://lemonade-server.ai/docs/embeddable/backends

Use this command to install the ROCm backend for llama.cpp on Windows if ROCm is supported by your system. The `stream: false` option indicates that the installation should not be streamed.

```bash
curl -X POST http://localhost:8000/v1/install \
  -H "Content-Type: application/json" \
  -d "{\"recipe\": \"llamacpp\", \"backend\": \"rocm\", \"stream\": false}"
```

--------------------------------

### Pi Configuration Files

Source: https://lemonade-server.ai/docs/integrations/pi

Example content for Pi's provider registration and default settings files.

```json
{
  "providers": {
    "Lemonade": {
      "baseUrl": "http://localhost:13305/v1",
      "api": "openai-completions",
      "apiKey": "lemonade",
      "models": [
        { "id": "Qwen3.5-35B-A3B-GGUF" }
      ]
    }
  }
}
```

```json
{
  "defaultProvider": "Lemonade",
  "defaultModel": "Qwen3.5-35B-A3B-GGUF"
}
```

--------------------------------

### Install .rpm package

Source: https://lemonade-server.ai/docs/dev/getting-started

Install the RPM package for specific Fedora versions.

```bash
# Fedora 43
sudo dnf install ./lemonade-server-<VERSION>-fc43.x86_64.rpm

# Fedora 44
sudo dnf install ./lemonade-server-<VERSION>-fc44.x86_64.rpm
```

--------------------------------

### Launch Open WebUI Server

Source: https://lemonade-server.ai/docs/integrations/open-webui

Start the Open WebUI HTTP server to begin the local interface session.

```bash
open-webui serve
```

--------------------------------

### Common Benchmark CLI Commands

Source: https://lemonade-server.ai/docs/guide/cli

Various command-line examples for running benchmarks with different models, backends, scenarios, and configurations.

```bash
# Benchmark with default scenarios and all installed backends
lemonade bench Qwen3-0.6B-GGUF

# Benchmark multiple models in one run. Note that Bonsai-4B is listed twice; it will only be tested once.
lemonade bench Bonsai-4B-gguf Gemma-4-E4B-it-GGUF Phi-4-mini-instruct-GGUF Qwen3-4B-GGUF Bonsai-4B-gguf Qwen3.5-4B-GGUF

# Benchmark specific backends and context sizes
lemonade bench --backend vulkan --backend cpu --ctx-size 4096 8192 Qwen3-0.6B-GGUF

# Run with custom scenarios
lemonade bench --scenario-file my-scenarios.json Qwen3-0.6B-GGUF

# Run only specific scenarios by name
lemonade bench --scenarios chat-short --scenarios code-debug Qwen3-0.6B-GGUF

# Run all scenarios in a category
lemonade bench --scenarios coding Qwen3-0.6B-GGUF

# Include long-context scenarios (excluded by default)
lemonade bench --scenarios all Qwen3-0.6B-GGUF

# Benchmark with custom backend arguments (e.g., different batch sizes)
lemonade bench --llamacpp-args "-b 1024" "-b 2048" Qwen3-0.6B-GGUF

# Save results as JSON
lemonade bench --json --output results.json Qwen3-0.6B-GGUF

# Compare against a previous run
lemonade bench --compare results.json Qwen3-0.6B-GGUF

# Run more measurement iterations for statistical significance
lemonade bench --runs 20 Qwen3-0.6B-GGUF
```

--------------------------------

### Configure CPU backend

Source: https://lemonade-server.ai/docs/guide/install/docker

Sets the backend to CPU in the configuration file and launches the container.

```json
{
  "llamacpp": {
    "backend": "cpu"
  }
}
```

```bash
docker run -d \
  --name lemonade-server \
  -p 13305:13305 \
  -v lemonade-cache:/opt/lemonade/.cache/huggingface \
  -v lemonade-llama:/opt/lemonade/llama \
  -v lemonade-recipe:/opt/lemonade/.cache/lemonade \
  ghcr.io/lemonade-sdk/lemonade-server:latest
```

--------------------------------

### Reranking Request Examples

Source: https://lemonade-server.ai/docs/api/llamacpp

Examples of how to send a reranking request using PowerShell and cURL.

```powershell
Invoke-WebRequest `
  -Uri "http://localhost:13305/v1/rerank" `
  -Method POST `
  -Headers @{ "Content-Type" = "application/json" } `
  -Body '{
    "model": "bge-reranker-v2-m3-GGUF",
    "query": "What is the capital of France?",
    "documents": [
      "Paris is the capital of France.",
      "Berlin is the capital of Germany.",
      "Madrid is the capital of Spain."
    ]
  }' -UseBasicParsing
```

```bash
curl -X POST http://localhost:13305/v1/rerank \
  -H "Content-Type: application/json" \
  -d '{
        "model": "bge-reranker-v2-m3-GGUF",
        "query": "What is the capital of France?",
        "documents": [
          "Paris is the capital of France.",
          "Berlin is the capital of Germany.",
          "Madrid is the capital of Spain."
        ]
      }'
```

--------------------------------

### Configure Backend Binary Versions

Source: https://lemonade-server.ai/docs/guide/configuration

Examples of setting backend binary paths and versions using the lemonade CLI. These commands apply changes live without requiring a service restart.

```bash
# Track upstream llama.cpp Vulkan releases (auto-resolve at lemond start)
lemonade config set llamacpp.vulkan_bin=latest

# Pin to a specific llama.cpp build
lemonade config set llamacpp.vulkan_bin=b8664

# Use your own llama.cpp build
lemonade config set llamacpp.vulkan_bin=/home/me/llama.cpp/build/bin

# Revert to the version lemonade ships
lemonade config set llamacpp.vulkan_bin=builtin
```

--------------------------------

### !help

Source: https://lemonade-server.ai/docs/integrations/mindcraft

Lists all available commands and their descriptions.

```APIDOC
## !help

### Description
Lists all available commands and their descriptions.
```

--------------------------------

### Configure System Backend

Source: https://lemonade-server.ai/docs/guide/configuration/llamacpp

Enable the system backend via configuration file or command line interface.

```json
"llamacpp": { "backend": "system" }
```

```bash
lemonade config set llamacpp.backend=system
```

--------------------------------

### Error Response Examples

Source: https://lemonade-server.ai/docs/api/openai

Example JSON error responses for common API failure scenarios.

```json
{"error": {"message": "Missing 'image' field (base64 encoded)", "type": "invalid_request_error"}}
```

```json
{"error": {"message": "Missing 'model' field", "type": "invalid_request_error"}}
```

```json
{"error": {"message": "Upscale model not found: bad-model", "type": "invalid_request_error"}}
```

```json
{"error": {"message": "ESRGAN upscale failed", "type": "server_error"}}
```

--------------------------------

### Install vLLM for ROCm

Source: https://lemonade-server.ai/docs/guide/configuration/vllm

Commands to install the vLLM backend for ROCm using the CLI or the HTTP API.

```bash
lemonade backends install vllm:rocm
```

```bash
curl -X POST http://localhost:13305/api/v1/install \
  -H 'Content-Type: application/json' \
  -d '{"recipe": "vllm", "backend": "rocm"}'
```

--------------------------------

### Install Claude Code

Source: https://lemonade-server.ai/docs/integrations/claude-code

Commands to install the Claude Code CLI tool using either curl or npm.

```bash
curl -fsSL https://claude.ai/install.sh | bash
```

```bash
npm install -g @anthropic-ai/claude-code
```

--------------------------------

### Build Windows MSI Installer

Source: https://lemonade-server.ai/docs/dev/getting-started

Executes the CMake build process to generate the WiX-based MSI installer for Windows.

```bash
cmake --build build --config Release --target wix_installers
```

--------------------------------

### Implement the Server Factory

Source: https://lemonade-server.ai/docs/dev/adding-a-backend

Implement the factory create function in the source file to instantiate the server with the provided BackendContext.

```cpp
#include "lemon/backends/myrecipe/myrecipe_server.h"
// … MyServer method definitions …

namespace lemon { namespace backends { namespace myrecipe {
std::unique_ptr<WrappedServer> create(const BackendContext& ctx) {
    return std::make_unique<MyServer>(ctx.log_level, ctx.model_manager, ctx.backend_manager);
}
}}}  // namespace lemon::backends::myrecipe
```

--------------------------------

### Lemonade Load Command Examples

Source: https://lemonade-server.ai/docs/guide/cli

Common usage scenarios for loading models with various configuration options.

```bash
# Load a model with default options
lemonade load Qwen3-0.6B-GGUF

# Load a model with custom context size
lemonade load Qwen3-0.6B-GGUF --ctx-size 8192

# Load a model and save options for future use
lemonade load Qwen3-0.6B-GGUF --ctx-size 4096 --save-options

# Load a llama.cpp model with custom backend
lemonade load Qwen3-0.6B-GGUF --llamacpp vulkan

# Load a llama.cpp model with custom arguments
lemonade load Qwen3-0.6B-GGUF --llamacpp-args "--flash-attn on --no-mmap"

# Load a model without merging global args (per-model args replace global entirely)
lemonade load Qwen3-0.6B-GGUF --no-merge-args --llamacpp-args "--flash-attn on"

# Load an image generation model with custom settings
lemonade load Z-Image-Turbo --sdcpp rocm --steps 8 --cfg-scale 1 --width 1024 --height 1024
```

--------------------------------

### Build and Development Commands

Source: https://lemonade-server.ai/docs/dev/app

Commands for installing dependencies, running the application in development mode, and creating production builds.

```bash
cd src/app

# Install webpack + Tauri CLI dependencies
npm ci

# Run in dev mode (opens a window, hot-reloads webpack)
npm run dev

# Production build (single binary, no OS bundles)
npm run build -- --no-bundle

# Production build with platform bundles (macOS .app, Linux .deb/.rpm, Windows MSI/NSIS)
npm run build
```

--------------------------------

### Configure Cloud Observability

Source: https://lemonade-server.ai/docs/guide/telemetry

CLI command to stream trace data directly to a cloud-hosted provider like Honeycomb.

```bash
lemonade config set telemetry.enabled=true \
                    telemetry.otlp.endpoint=https://api.honeycomb.io/v1/traces \
                    telemetry.otlp.semantics='["otel_genai"]' \
                    telemetry.otlp.headers='{"x-honeycomb-team": "YOUR_API_KEY"}'
```

--------------------------------

### Install Lemonade RPM on Fedora 44

Source: https://lemonade-server.ai/docs/guide/install/fedora

Installs the specific RPM package for Fedora 44 using the dnf package manager.

```bash
sudo dnf install ./lemonade-server-*-fc44.x86_64.rpm
```

--------------------------------

### Install Lemonade RPM on Fedora 43

Source: https://lemonade-server.ai/docs/guide/install/fedora

Installs the specific RPM package for Fedora 43 using the dnf package manager.

```bash
sudo dnf install ./lemonade-server-*-fc43.x86_64.rpm
```

--------------------------------

### Installing Cloud Providers

Source: https://lemonade-server.ai/docs/dev/router-policy

Configure cloud model providers before registering collections to ensure proper resolution.

```bash
export LEMONADE_FIREWORKS_API_KEY=fw-XXXX
lemonade cloud install fireworks --base-url https://api.fireworks.ai/inference/v1
lemonade list | grep fireworks     # e.g. fireworks.kimi-k2p5
```

--------------------------------

### Retrieve Runtime Configuration

Source: https://lemonade-server.ai/docs/embeddable/runtime

Fetches the current server-level and recipe configuration snapshot.

```cmd
curl http://localhost:8000/internal/config
```

```bash
curl http://localhost:8000/internal/config
```

--------------------------------

### Install Lemonade Desktop App via Snap

Source: https://lemonade-server.ai/docs/guide/install/ubuntu

Install the fully sandboxed Lemonade desktop application using snap. This provides a self-contained desktop experience.

```bash
sudo snap install lemonade

```

--------------------------------

### Launch options for chat

Source: https://lemonade-server.ai/docs/guide/cli-chat

Alternative methods to start the chat REPL, including specifying a model or using the run command with the chat flag.

```bash
# Open the REPL, optionally with a model
lemonade chat
lemonade chat Qwen3-1.7B-Hybrid

# Or piggyback on the existing run command
lemonade run Qwen3-1.7B-Hybrid --chat-cli
```

--------------------------------

### Install Lemonade Server via PPA

Source: https://lemonade-server.ai/docs/guide/install/ubuntu

Add the stable PPA for Lemonade and install the server package. This method is suitable for users who prefer managing packages through apt.

```bash
sudo add-apt-repository ppa:lemonade-team/stable
sudo apt install lemonade-server

```

--------------------------------

### Upscale Image Workflow

Source: https://lemonade-server.ai/docs/api/openai

Examples demonstrating the generation of an image followed by an upscale request using Bash, PowerShell, and Python.

```bash
# Step 1: Generate an image and save the base64 response
RESPONSE=$(curl -s -X POST http://localhost:13305/v1/images/generations \
  -H "Content-Type: application/json" \
  -d '{
        "model": "SD-Turbo",
        "prompt": "A serene mountain landscape at sunset",
        "size": "512x512",
        "steps": 4,
        "response_format": "b64_json"
      }')

# Step 2: Build the upscale JSON payload and pipe it to curl via stdin
# (base64 images are too large for command-line interpolation)
echo "$RESPONSE" | python3 -c "
import sys, json
b64 = json.load(sys.stdin)['data'][0]['b64_json']
print(json.dumps({'image': b64, 'model': 'RealESRGAN-x4plus'}))
" | curl -X POST http://localhost:13305/v1/images/upscale \
  -H "Content-Type: application/json" \
  -d @-
```

```powershell
# Step 1: Generate an image
$genResponse = Invoke-WebRequest `
  -Uri "http://localhost:13305/v1/images/generations" `
  -Method POST `
  -Headers @{ "Content-Type" = "application/json" } `
  -Body '{
    "model": "SD-Turbo",
    "prompt": "A serene mountain landscape at sunset",
    "size": "512x512",
    "steps": 4,
    "response_format": "b64_json"
  }'

# Step 2: Extract the base64 image
$imageB64 = ($genResponse.Content | ConvertFrom-Json).data[0].b64_json

# Step 3: Upscale the image with Real-ESRGAN
$body = @{ image = $imageB64; model = "RealESRGAN-x4plus" } | ConvertTo-Json
Invoke-WebRequest `
  -Uri "http://localhost:13305/v1/images/upscale" `
  -Method POST `
  -Headers @{ "Content-Type" = "application/json" } `
  -Body $body
```

```python
import requests
import base64

BASE_URL = "http://localhost:13305/api/v1"

# Step 1: Generate an image
gen_response = requests.post(f"{BASE_URL}/images/generations", json={
    "model": "SD-Turbo",
    "prompt": "A serene mountain landscape at sunset",
    "size": "512x512",
    "steps": 4,
    "response_format": "b64_json",
})
image_b64 = gen_response.json()["data"][0]["b64_json"]

# Step 2: Upscale the image with Real-ESRGAN (512x512 -> 2048x2048)
upscale_response = requests.post(f"{BASE_URL}/images/upscale", json={
    "image": image_b64,
    "model": "RealESRGAN-x4plus",
})

# Step 3: Save the upscaled image to a file
upscaled_b64 = upscale_response.json()["data"][0]["b64_json"]
with open("upscaled.png", "wb") as f:
    f.write(base64.b64decode(upscaled_b64))
```

--------------------------------

### List models command

Source: https://lemonade-server.ai/docs/guide/cli

Displays all models, optionally filtered by the --downloaded flag.

```bash
lemonade list [options]
```

--------------------------------

### Install Lemonade Desktop App via Apt

Source: https://lemonade-server.ai/docs/guide/install/ubuntu

Install the Lemonade desktop application using apt, which includes the web UI. This package is useful for users who want a desktop integration.

```bash
sudo apt install lemonade-desktop

```

--------------------------------

### !startConversation

Source: https://lemonade-server.ai/docs/integrations/mindcraft

Start a conversation with a player.

```APIDOC
## !startConversation

### Description
Start a conversation with a player. Use for bots only.

### Parameters
- **player_name** (string) - The name of the player to send the message to.
- **message** (string) - The message to send.
```

--------------------------------

### Run Frontend Development Mode

Source: https://lemonade-server.ai/docs/dev/getting-started

Starts the Tauri development server with hot reloading for UI iteration.

```bash
cd src/app
npm run dev
```
Tokens:
Raw