Claude Code makes local LLMs 90% slower(unsloth.ai)

https://unsloth.ai/docs/basics/claude-code

1 comments

The relevant part is that Claude Code adds an Attribution header, which invalidates the KV Cache for local models. The fix proposed here is to set the following in ~/.claude/settings.json when using Claude Code against a local model:

    "env": {
      ...,
      "CLAUDE_CODE_ATTRIBUTION_HEADER" : "0"
    },