Skip to main content

vai nano

Set up and manage local voyage-4-nano inference in vai.

Synopsis​

vai nano <subcommand>

Description​

vai nano is the control surface for local inference in vai v1.31.0. It manages the Python environment, model download, readiness checks, and cache lifecycle for voyage-4-nano.

This command is the entry point for the local-first workflow:

vai nano setup
vai embed "What is vector search?" --local

Subcommands​

SubcommandDescription
setupInstall dependencies, create the local environment, and download the model
statusShow whether local inference is ready
testRun a smoke test against the local model
infoShow model, cache, and environment details
clear-cacheRemove cached model files

Examples​

Set up local inference​

vai nano setup

Check readiness​

vai nano status

Smoke test the local model​

vai nano test

Show environment details​

vai nano info

Clear cached model files​

vai nano clear-cache

Notes​

  • vai nano setup is the one-time setup step for local voyage-4-nano inference
  • local inference uses a lightweight Python bridge, so Python 3.9+ is required
  • once setup is complete, use --local on supported commands such as vai embed and vai pipeline