Running Google Gemma 4 Locally with LM Studio's New Headless CLI and Claude Code

Using Claude Code seems like a popular frontend currently, I wonder how long until Anthropic releases an update to make it a little to a lot less turn-key? They've been very clear that they aren't exactly champions of this stuff being used outside of very specific ways.

Just FYI, MoE doesn't really save (V)RAM. You still need all weights loaded in memory, it just means you consult less per forward pass. So it improves tok/s but not vram usage.

Here is how I set up Gemma 4 26B for local inference on macOS that can be used with Claude Code.

So wait what is the interaction between Gemma and Claude?

  ollama launch claude --model gemma4:26b