Try prompt compression on real context.

Paste a support ticket, RAG chunk, or agent trace. Adola returns the smaller context it would pass to your model, plus a receipt for what changed.

Demo API

Call the limited demo endpoint directly.

The public demo route is capped for quick testing. Create a workspace when you are ready for production keys, higher limits, and metered receipts.

Send query and context Receive smaller text Inspect token receipt

curl -s https://api.adola.app/v1/demo/compress \
  -H 'content-type: application/json' \
  --data '{
    "model": "rose-1",
    "query": "What should finance approve next?",
    "input": "Long RAG or agent context...",
    "compression": { "target_ratio": 0.35 }
  }'

Open API docs