yourai build
Shows the model license and asks you to accept it, then downloads the engine and the model
into yourai_module/ (gitignored, like node_modules) and writes
yourai.lock (safe to commit). Interrupted downloads resume. Each file's
SHA-256 is checked. In scripts, add --accept-license.
Flag combos:
yourai build— use the shared~/.youraicache if present, else download intoyourai_module/onlyyourai build -s— same, and also write new downloads into~/.youraiyourai build -f— always download online intoyourai_module/yourai build -f -s— download online and update the saved~/.youraicopy
Delete yourai_module/ anytime to free disk; run yourai build again to restore it.
Example
$ cd study-buddy && yourai build
Model license
Qwen2.5 0.5B Instruct is distributed under: Apache-2.0
? Accept this license and download the model? Yes
✔ engine llama-b12015-bin-macos-arm64.tar.gz downloaded and verified
✔ engine llama.cpp b12015 (macos-aarch64-metal) installed
✔ model Qwen2.5 0.5B Instruct downloaded and verified
✔ study-buddy is built
Next: yourai start
Questions or problems? support@bdtren.com · Contact page.