big prompt bad.
small prompt good.

grunt rewrites bloated prompts into the smallest
fragment that
still means the same thing. 40-60% fewer tokens,
zero
loss of technical meaning.

try it now →
grunt.compress(prompt) SAMPLE:
verboseramblingover-polite
Before45 tok
"Could you please help me write a function that checks whether a given number is prime? I would really appreciate it if you could also include some comments explaining how it works."
After • Full11 tok
~50%
avg token reduction
4
compression levels
10/day
free tier
real compression. real numbers.

same tool, same level (full), three different input types.

Pure prose71% saved
before · 292 chars

I wanted to reach out to ask whether you would be able to help me understand the best approach to take when it comes to building a content strategy for a new product that is just getting started and does not yet have an established audience or any significant brand recognition in the market.

after · 81 chars

How to build content strategy for new product with no audience/brand recognition?

~52 tokens saved$0.156 / 1k calls

Code blocks, URLs, and file paths are preserved byte-for-byte — they can't be compressed without breaking them. Savings are highest on pure prose, lower when technical regions dominate the input.

coming soon
grunt APIcoming soon

compress prompts programmatically. REST endpoint, API key auth, all 4 levels.

POST /api/v1/compress
Authorization: Bearer grk_…
pricing
Free
$0
forever
  • ◆ 10 compressions / day
  • ◆ lite + full levels
  • ◆ token savings counter
get started
Prorecommended
$9
per month
  • ◆ unlimited compressions
  • ◆ all 4 levels incl. wenyan
  • ◆ prompt history library
upgrade to pro