Terse — Token Optimizer for LLMs

About

Terse compresses your AI prompts in real-time, saving tokens and reducing API costs — without losing meaning.

How it works:
Terse runs a 20+ technique optimization engine directly in your browser. It removes filler words, politeness padding, hedging, redundant phrases, and compresses whitespace — all while preserving the meaning of your prompt. Your text never leaves your browser unless you choose to sign in.

Three optimization modes:
- Soft — Typo correction + whitespace compression. 100% meaning preserved.
- Normal — Removes fillers, politeness, hedging, meta-language, shortens phrases.
- Aggressive — Maximum compression. Abbreviations, article removal, telegraph style.

Works with every AI chat:
Terse works on any AI chat or agent — if you can type into it, Terse can optimize it.

Features:
- One-click Capture & Replace — reads your prompt, optimizes it, writes it back
- Live monitoring — auto-optimizes as you type (optional)
- Manual paste mode — paste any text to optimize
- Token counter — see exactly how many tokens you save
- Technique tags — see which optimizations were applied
- 8 beautiful themes (Lime, Lavender, Coral, Teal, Midnight, Rose, Sage, Sand)
- Keyboard shortcuts (Cmd+Shift+T to optimize, Cmd+Shift+R to replace)
- Undo — instantly revert any replacement
- All optimization runs locally in your browser — fast and private

Saves real money:
If you use AI APIs, every token counts. Terse typically saves 15–40% of tokens per prompt, which adds up fast when you're paying per million tokens.

Privacy first:
All optimization happens locally in your browser. No text is sent to any server. Optional sign-in enables usage sync and subscription features.