MonkeyCode
A self-hosted AI coding platform for teams, with offline private deployment, server-side task environments and GLM, Qwen and DeepSeek support.
MonkeyCode is an AI coding platform for engineering teams that want to delegate development work and manage it centrally. It supports private offline deployment inside a team's network, where data can stay local.
- Requirements and tasks Teams can manage requirements and specifications alongside AI tasks, handing work to agents for implementation and validation.
- Development environments Each task runs in a server-side environment for building, testing and previewing code.
- Code review and collaboration Automated pull request and merge request reviews sit alongside shared project work and file management.
- Model selection Integrations include GLM, Kimi, MiniMax, Qwen and DeepSeek, with manual selection or switching according to the task.
- Mobile access Native iOS and Android support synchronizes desktop and mobile task data so you can check work and files while agents continue running.
MonkeyCode is open source under AGPL-3.0. Its hosted browser service requires an account and provides managed development environments and built-in models; tasks on that service run in the cloud. The platform doesn't provide a local IDE, local CLI or code completion.
For self-hosting, the recommended console resources are 2 CPU cores, 4 GB RAM and 40 GB storage. The development environment host has a separate recommendation of 8 CPU cores, 16 GB RAM and 100 GB storage.
