token compressor for Zed

io.github.base76-research-lab/token-compressor

Compress prompts 40-60% using local LLM + embedding validation. Preserves all conditionals.

client:Zed transport:stdio runtime:pypi

Install token compressor in Zed

~/.config/zed/settings.json · windows: %APPDATA%\Zed\settings.json

{
  "context_servers": {
    "token-compressor": {
      "command": "uvx",
      "args": [
        "token-compressor-mcp"
      ]
    }
  }
}

Merge this into your existing settings object — do not replace the file. Zed docs