Give it a task. Let it get to work.
Set the goal and let Ncurator handle the steps.
Ncurator, your AI work partner
Give Ncurator a task. It can use your local knowledge base, search the web, and work with Word, PDF, Excel and PowerPoint.
AI with your local knowledge
Ncurator uses your local knowledge base to find relevant answers and work with your own material.
Your models. Your costs.
The app is free. You pay your AI provider for the tokens you use, with no mandatory credit bundles. A local knowledge base helps reduce token usage.
An AI work partner that uses fewer tokens
Ncurator reuses cached input across tasks, so repeated content can be billed at the lower cache-hit rate.
Cache hit rate for repeated tasks
Most input tokens in repeated tasks can use the lower cache-hit rate.
A lower price for cached input
In this DeepSeek pricing example, cached input costs one thirtieth of uncached input.
About 90% less per task
A task priced at $4.49 would cost about $0.45 with the cache hit rate shown here.
$4.49 × (93% ÷ 30 + 7%) ≈ $0.45What Ncurator can do
08Read browser bookmarksExtension required
After an edit, use Preview to see the result or Review to check what the AI changed.
This week's business report
The user's request@standup-notes.txt @customer-risks.txt @weekly-numbers.md Write this week's business report for management, covering progress, numbers, risks and next steps.