Rtk
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
Last verified:
What is Rtk?
RTK (Rust Token Killer) is an open-source CLI proxy tool designed for developers using LLM-powered AI coding assistants. It sits between your AI assistant and development tools, filtering and compressing command output before it reaches the LLM's context window. RTK removes boilerplate, progress bars, and noise from CLI outputs while preserving meaningful content like test failures, error messages, diffs, and stack traces. The result is 60-90% fewer tokens consumed per command without changing how you work.
Key features include 89% average noise reduction measured across 2,900+ real-world dev commands, support for 30+ commands across major ecosystems (Git, Cargo/Rust, JavaScript, Python, Go, Ruby, .NET, Docker/Kubernetes), and zero configuration required. RTK is written in Rust as a single binary with no dependencies, is completely free under the MIT license, and includes real-time token savings analytics via the rtk gain command. It works transparently with Claude Code, Cursor, Aider, Gemini CLI, OpenAI Codex, Cline, Windsurf, and GitHub Copilot.
RTK is ideal for developers using AI coding assistants who want longer coding sessions, reduced API costs, and better AI reasoning quality. By reducing context pollution, it enables 3x longer sessions and can cut API bills by ~70% for pay-per-token setups. A typical team of 10 can save approximately $1,750/month in wasted token costs. The tool is particularly valuable for those on flat-rate plans hitting rate limits faster than expected or teams using API-based tools like Aider and Gemini CLI.
Rtk pricing
Pricing model: Freemium
RTK is completely free and open source under the MIT license with no usage limits, no API keys required, no telemetry, and no accounts needed. Source code is available on GitHub at github.com/rtk-ai/rtk. RTK Cloud (not yet launched, on waitlist) will offer team features starting at $15 per developer per month, including token analytics dashboards per dev/project/tool, team savings reports, rate limit alerts, and enterprise controls like SSO and audit logs.
Rtk pros
- 89% average noise reduction across 2,900+ real commands
- 60-90% token savings per command
- Completely free and open source under MIT license
- Single Rust binary with zero dependencies
- Zero configuration required - works out of the box
- Works with all major AI coding assistants
- Supports 30+ CLI commands across ecosystems
- 3x longer AI coding sessions
- Reduces API costs by approximately 70%
- Real-time token savings tracking via rtk gain command
- Preserves test failures and error messages fully
- Intercepts commands automatically via Bash hook
- No API keys or accounts required
- No telemetry or data collection
- Cross-platform support for macOS, Linux, Windows
- 3x better context quality for AI reasoning
- Immediate installation via curl or Homebrew
- Tracks savings by command with detailed analytics
- Compresses git diff by up to 94%
- Compresses cargo test by 91.8%
- Compresses git status by 80.8%
- Compresses file reads by 60-95%
- Active GitHub community with 54.8k stars
- Upcoming RTK Cloud for team cost visibility
Rtk cons
- Still in early stages with RTK Cloud not yet launched
- Only 0 teams currently on RTK Cloud waitlist
- Primarily focused on CLI-based AI assistants
- May not help with non-terminal AI interactions
- grep compression only 49.5% (lowest among major commands)
- find compression only 78.3% (lower than average)
- Requires terminal/AI assistant integration to work
- Does not reduce model input tokens from file content directly
- Windows users must extract zip and use CLI not double-click
- Learning curve for understanding filter strategies
- No official GUI interface available
- Limited to developer/technical users
- No enterprise SSO or compliance features yet
- Advanced team features only coming in RTK Cloud
- Does not work with completely GUI-based AI tools
Frequently asked questions about Rtk
What is RTK (Rust Token Killer)?
RTK is an open-source CLI tool that compresses command outputs before they reach the AI context window. It reduces token usage by 60-90% with zero configuration changes, enabling longer AI coding sessions and lower API costs. RTK is written in Rust, MIT licensed, and works transparently with Claude Code, Cursor, and any terminal-based AI assistant.
How many tokens does RTK actually save?
Based on measurements across 2,900+ real-world commands, RTK removes an average of 89% of CLI output noise. Command-level savings include cargo test at 91.8% savings, git status at 80.8%, find at 78.3%, and grep at 49.5%. A developer who ran 15,720 commands saved 138 million tokens over several weeks, tracked live via rtk gain.
Is RTK free? Are there usage limits?
RTK is completely free. It is open source under the MIT license, with source code available on GitHub at github.com/rtk-ai/rtk. There are no usage limits, no API keys required, no telemetry, and no accounts. RTK Cloud (waitlist) will offer additional team features.
Which AI coding tools and commands does RTK support?
RTK works with Claude Code (Anthropic), Cursor, Gemini CLI, Aider, and any AI assistant that reads terminal output. It supports all major CLI commands including cargo test, pytest, go test, git diff, git status, git log, grep, find, ls, pnpm list, tsc, eslint, prisma, docker, kubectl, and more across Git, Cargo/Rust, JavaScript, Python, Go, Ruby, .NET, and Docker/Kubernetes ecosystems.
How does RTK work without changing my workflow?
Running rtk init --global installs a PreToolUse hook in Claude Code that automatically rewrites Bash commands to rtk equivalents at the proxy layer. You continue using your normal commands. RTK intercepts and compresses the output before it enters the context window, with no manual changes to prompts or workflows needed.
Does RTK affect code quality or AI reasoning accuracy?
No. RTK removes verbose boilerplate and repetitive output noise, not meaningful content. Test failures, error messages, diffs, and stack traces are preserved in full. The AI receives the same essential information with 89% less noise, which typically improves reasoning quality by reducing context pollution.
How do I install RTK?
Install via one-liner for Linux/macOS: curl -fsSL https://raw.githubusercontent.com/rtk-ai/rtk/refs/heads/master/install.sh | sh, or via Homebrew: brew install rtk. Pre-built binaries are available for macOS, Linux, and Windows from GitHub releases. After installation, activate with rtk init --claude-code or rtk init --global for global hook installation.
How do I measure my token savings with RTK?
Simply run rtk gain to see real-time token savings analytics. The command displays total commands processed, input tokens, output tokens, tokens saved with percentage, and a breakdown by command. A happy developer's feedback showed 15,720 commands processed with 138M tokens saved after several weeks of daily use, achieving 88.9% efficiency.
What is the difference between RTK and other token savers?
RTK is a CLI proxy that compresses output at the proxy layer before it reaches any AI context window, so any tool that reads terminal output gets cleaner context. Unlike tool-specific solutions, RTK works with every AI coding tool out of the box with zero config. It is written in Rust as a single binary with no dependencies, is completely free and open source, and achieves 89% average noise reduction.
What is RTK Cloud and when will it launch?
RTK Cloud is an upcoming team feature offering visibility and control over team AI coding costs. It includes token analytics dashboards per dev/project/tool, team savings reports showing amounts like $4,200 saved in a month, rate limit alerts with monitoring and smart notifications, and enterprise controls including SSO, audit logs, and compliance. It is currently on waitlist with 0 teams signed up, launching at $15 per developer per month for teams while remaining free for open-source.