Kode's Docker deployment requires special attention to data persistence configuration:
- infrastructure
docker build --no-cache -t kode .Use no-cache to keep your dependencies up to date - critical mount point
Three paths must be mapped:-v $(pwd):/workspace: Mount the current project directory-v ~/.kode:/root/.kode: Persistent Model Cache-v ~/.kode.json:/root/.kode.json: Preserve configurations such as API keys
- Full Startup Example
docker run -it --rm -v /path/to/project:/workspace -v ~/kode-data:/root/.kode -w /workspace kode --safe
Production environment recommendations:
- Using named volume instead of bind mount to improve performance
- Predefine memory limits for each model in docker-compose
- pass (a bill or inspection etc)
--model-queue-limitParameters control the number of concurrent requests
This answer comes from the articleKode: Claude Code Open Source Optimized VersionThe
May not be reproduced without permission:AI productivity tools " How to deploy Kode via Docker and keep the configuration persistent?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil