🍺 BREW Explorer

← all formulae

rtk

brew install rtk v0.42.4 Apache-2.0

CLI proxy that reduces LLM token consumption by caching and filtering redundant output before sending to Claude or other AI models.

Why you might care

If you're chaining shell commands into Claude Code or similar AI coding assistants, RTK caches results and strips boilerplate to cut token overhead by 60–90%, letting you run longer sessions without hitting rate limits. It's a transparent proxy that requires zero config and works with existing workflows.

Categories

proxy cli-tool networking

Alternatives

llm aider
25.0k
30-day installs · #208
66.8k
90-day · #230
73.7k
365-day · #521

Build dependencies

rust

Links

Blurb generated by claude-haiku-4-5 on today.

Raw metadata 
{
  "aliases": [],
  "alternatives": [
    "llm",
    "aider"
  ],
  "build_dependencies": [
    "rust"
  ],
  "categories": [
    "proxy",
    "cli-tool",
    "networking"
  ],
  "caveats": null,
  "conflicts_with": [],
  "dependencies": [],
  "deprecated": 0,
  "deprecation_reason": null,
  "desc": "CLI proxy to minimize LLM token consumption",
  "disable_reason": null,
  "disabled": 0,
  "enrichment_fetched_at": "2026-06-20T23:40:44+00:00",
  "first_seen": "2026-06-20T23:34:18+00:00",
  "full_name": "rtk",
  "github_default_branch": null,
  "github_last_commit_at": null,
  "github_readme_excerpt": null,
  "github_repo": null,
  "github_stars": null,
  "github_topics": [],
  "homepage": "https://www.rtk-ai.app/",
  "homepage_og_description": "RTK reduces Claude Code token usage by 60-90% \u2014 measured across 2,900+ real commands. 89% avg noise removed, 3x longer sessions, zero config. Apache 2.0 licensed.",
  "homepage_og_image": "https://www.rtk-ai.app/og-image.png",
  "homepage_title": "RTK \u2014 Rust Token Killer",
  "installs_30d": 24962,
  "installs_365d": 73684,
  "installs_90d": 66830,
  "keg_only": 0,
  "keg_only_reason": null,
  "last_seen": "2026-06-20T23:34:18+00:00",
  "license": "Apache-2.0",
  "llm_generated_at": "2026-06-20T23:43:24+00:00",
  "llm_model": "claude-haiku-4-5",
  "name": "rtk",
  "oldnames": [],
  "one_liner": "CLI proxy that reduces LLM token consumption by caching and filtering redundant output before sending to Claude or other AI models.",
  "optional_dependencies": [],
  "rank_30d": 208,
  "rank_365d": 521,
  "rank_90d": 230,
  "raw_hash": "aa3c5d307a950037",
  "recommended_dependencies": [],
  "revision": 0,
  "ruby_source_path": "Formula/r/rtk.rb",
  "tap": "homebrew/core",
  "test_dependencies": [],
  "uses_from_macos": [],
  "version_head": "HEAD",
  "version_stable": "0.42.4",
  "versioned_formulae": [],
  "why_use_this": "If you\u0027re chaining shell commands into Claude Code or similar AI coding assistants, RTK caches results and strips boilerplate to cut token overhead by 60\u201390%, letting you run longer sessions without hitting rate limits. It\u0027s a transparent proxy that requires zero config and works with existing workflows."
}