Terse auto-appears when you focus Cursor, compresses your prompts in real time, and shows you the savings before you send. 40–70% fewer tokens per message.
Terse reads from Cursor's Chat and Composer panels via macOS Accessibility APIs, compresses on-device, and shows the result before you hit send.
When you click into Cursor's Chat or Composer, Terse auto-appears at the top of your screen — without stealing focus or interrupting your flow.
Auto-detectPhrases like "Could you possibly help me…", "I was hoping you could…", and "It would be great if…" add tokens without adding meaning. Terse removes them automatically.
Normal ModeVariable names, file paths, and backtick code blocks are always excluded from compression. Only natural language prose is touched — never your code.
Always SafeEnable Auto in the Terse popup to automatically replace every Cursor prompt with its optimized version before you send. Zero extra clicks.
One-ClickSee the before and after token count for every prompt in real time. Know exactly how much you're saving on each Cursor request.
Real-timeSoft (typo fix only), Normal (filler removal), and Aggressive (maximum compression). Switch modes per-prompt depending on how much context precision you need.
FlexibleCursor's cost isn't just your prompts — it's everything Cursor adds automatically before your words even reach the model.
Everything about using Terse with Cursor.
Terse supports every major AI coding tool and browser-based AI assistant.
30-day free trial. No credit card required until your trial ends. Cancel anytime.