Project
- Stars
- Started
- Issues
- License
reComputer RK LLM
Self-hostedRun local language models on the reComputer
reComputer RK LLM runs supported language and vision-language models on the board’s Rockchip NPU. It packages the runtime and a local server in Docker, with an interactive chat mode and API endpoints that other apps can use.
Use the RK3588 image and a model that fits the board’s available memory. Start with a small model on an 8 GB board. The server needs NPU device access and has no built-in login or TLS, so keep it on a trusted network. Rockchip runtime binaries and model weights have separate terms.
- Run AI models locally
Feedback
How did it go? Tips, problems, questions…
Feedback
How did it go? Tips, problems, questions…