Hey Product Hunt community! 👋 I am super excited to launch ShrinkToken today! The Inspiration: Like many of you using ChatGPT, Claude, and Gemini every day for massive document analysis or coding tasks, I kept running into two annoying problems: escalating API token costs and hitting rigid context window character limits. When I analyzed my chat history, I realized almost 40-50% of my token consumption was being wasted on conversational pleasantries ("please", "would you kindly", "I appreciate"), repetitive intensifiers, and redundant syntactic formatting that AI models don't even need to execute instructions accurately. Why ShrinkToken is different (Zero-Trust Privacy): Recent reports about third-party AI helper apps intercepting and logging customer prompts made me realize that transmitting proprietary code through an external online "prompt optimizer server" is a major security risk. I engineered ShrinkToken to execute 100% offline inside your local browser DOM. Zero bytes are ever sent over HTTP to external servers—your conversations and corporate code never leave your machine! ✨ What is available to use right now today: - Persistent Chrome Side Panel: Opens a sleek, dark-mode cyberpunk side panel right alongside your active ChatGPT or Claude tabs. Just paste your verbose instructions, watch the engine strip out the filler in milliseconds, and copy the streamlined prompt with up to 50% fewer tokens! - Real-Time Telemetry Dashboard: Tracks your live token reductions and maintains an offline lifetime savings counter across all your AI sessions. - Standalone Instant Web App: Working on a locked-down work laptop or browsing on mobile? Open shrinktoken.netlify.app in any browser (Safari, Firefox, Edge) to compress prompts instantly without installing an extension! Everything is fully open-source on GitHub for anyone to audit or contribute to. I would love to hear your honest feedback on our linguistic pruning algorithms! How many tokens does it save you today? Let's discuss below! ⚡

ShrinkToken

Offline AI prompt compressor to save up to 50% on tokens

Offline AI prompt compressor to save up to 50% on tokens