Launched this week
TerraWatch AI Prompt Compressor
Auto-compress your AI prompts and cut token costs by 40%
2 followers
Auto-compress your AI prompts and cut token costs by 40%
2 followers
TerraWatch compresses your prompts before they reach the AI — automatically. No extra clicks, no workflow changes. Just type and send like normal, and a toast shows you how many tokens were saved. Works as a Chrome extension (auto-compresses on submit) and a CLI for developers piping prompts into the API. Free. No account needed. Works on the major AI chat tools.
Launch Team
AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted
hey PH 👋
i built TerraWatch after noticing my own prompts were way longer than they needed to be. redundant context, filler phrases, restating things the model already knew. the tokens were just wasted.
started as a manual tool you’d click a button to compress before sending. people installed it and never used it. the extra step killed it.
so i rebuilt it to intercept the send button automatically. now it just happens. you type, you hit send, a small toast tells you how many tokens were saved. that’s it.
free chrome extension + CLI for devs who want to pipe it into their API workflow.
would love any feedback especially from people building on top of LLMs who are watching costs scale.