Pre-LLM token compressor for logs, JSON, and tool outputs, saving 20-95% tokens.
Worth a timeboxed spike before you bet on it.
Small teams using LLM agents (Claude Code, Cursor) and hitting token limits or high costs.
Manual prompt trimming or paying for larger context windows.
Compression may lose nuance in critical logs; needs careful testing per use case.
Install via pip, run the proxy on a sample JSON log file and compare output quality.
The numbers
Maintainers describe it as: “Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.”
headroom, in short
- Should a small team use headroom?
- Pilot it. Worth a timeboxed spike before you bet on it. Small teams using LLM agents (Claude Code, Cursor) and hitting token limits or high costs.
- What does headroom actually do?
- Pre-LLM token compressor for logs, JSON, and tool outputs, saving 20-95% tokens.
- What does headroom replace?
- Manual prompt trimming or paying for larger context windows.
- What is the downside of headroom?
- Compression may lose nuance in critical logs; needs careful testing per use case.
- Can headroom be used in a commercial product?
- Its licence is Apache-2.0, which is permissive and generally fine for commercial use. Confirm against the LICENSE file in the repository.
Weighed against
Turns any website into clean markdown an LLM can read, including pages that need JavaScript to render.
A Python scraping library that adapts when a site changes its markup, instead of breaking silently.
AI-powered desktop assistant
CLI tool for AI agents to search internet platforms