Media
2 items
Description
One-click text compression for AI prompts. Strips scaffolding, hedges, and bloat. 100% local, no API calls. Trim the bloat before you hit send. LLM responses are full of scaffolding ("Here's what we'll do..."), hedges ("It's worth noting..."), recap ("In summary..."), and corporate softeners ("In order to..."). When you forward that text to another LLM as context, you pay token cost twice — once to receive it, once to re-send it. TrimText Quick is the keyboard-shortcut companion to trimtext.dev. Open the popup, paste the bloated text, hit Compress, copy the result. The popup reports word-count savings so you can see how much context bloat was removed before sending. HOW IT WORKS • Deterministic regex rules — no LLM, no API call, no waiting • Strips opening scaffolds, recap framing, hedges, and meta-commentary • Substitutes verbose phrasing ("in order to" to "to", "due to the fact that" to "because") • Preserves markdown structure: headers, bullets, code blocks, numbered lists • Reports word-count savings inline USE CASES • Cleaning up an LLM response before pasting it into a follow-up prompt • Compressing meeting notes or transcripts before AI summarization • Tightening blog drafts and emails (the rules are general-purpose, not AI-specific) • Pre-processing context for long-context sessions PRIVACY • 100% local — every keystroke and every result stays in your browser • Zero network requests, zero analytics, zero accounts • Uninstall removes everything KEYBOARD SHORTCUT • Cmd+Shift+Y (Mac) / Alt+Shift+T (Windows/Linux) opens the popup • Cmd+Enter inside the popup runs Compress Prefer the web version? trimtext.dev runs the same compression rules in a full-page editor with a visible edit audit. More free tools: PasteDrop, TypingReps, and the Refresh All Tabs / TackNote / Site Bulletin extensions.