A VS Code extension that downloads, hosts, and runs the Lemonade Server embeddable binary (lemond) locally, enabling AI chat with local LLMs directly in VS Code.
⚠️ Early stage: The 0.0.1 release focuses primarily on setup: downloading and hosting the binary, plus basic server and model management.
More functionality is planned for upcoming releases.
Testing is still limited, and development has so far been done only on Windows.
Other platforms and edge cases may not work as expected. Feedback and issue reports are welcome.
Features
Managing/downloading lemonade models
Chatting integration
Chat Commands
The @lemon chat participant supports the following slash commands:
/fix - Generate a fix for the selected code
/explain - Explain the selected code
Configuration
Setting
Default
Description
lemon.chatModel
""
Model to use for chat (leave empty to be prompted)
lemon.customServerUrl
(unset)
Custom Lemonade Server URL used when lemon.targetServer is "custom" (e.g., http://localhost:13305)
lemon.embeddedPort
8000
Port for the embedded lemon server
lemon.maxLoadedModels
(unset)
Maximum number of loaded models. Use -1 for unlimited.
lemon.standalonePort
13305
Port for the standalone Lemonade Server
lemon.targetServer
"standalone"
Which Lemonade Server to use: standalone, embedded, or custom
Server Mode Selection
You can choose which Lemonade server the extension uses by setting lemon.targetServer (Lemon: Select Server lets you pick from the command palette, which updates this setting automatically):
Value
Behavior
standalone (default)
Connects to an existing system-installed standalone Lemonade Server.
