Docs
K‑Dense Web does a lot, and almost none of it is discoverable by poking at the interface. These are the pages worth reading first.
Guide to prompting K-Dense Web
The six things worth specifying before you hit go: objective, data source, deliverables, method preferences, audience, and context. With before-and-after examples. This is the highest-leverage thing to read.
ReadEffort levels
Instant, Standard, and Pro — what each one does, whether it can revise its own plan, which validation passes run, and how to pick. The most common cause of disappointing output is running complex work at the wrong level.
ReadWebMCP for AI agents
K-Dense Web supports WebMCP, so an AI agent in your browser can start research, follow the run, and read the report through typed tools. What each page offers, what stays with you, and how to turn it on.
ReadSecurity & data handling
What we do with your data, what we don't, how deletion works, where runs can execute, and our current compliance position stated plainly.
ReadVideo tutorials
Step-by-step walkthroughs of real sessions, from setting up a run to pulling the deliverables out at the end.
ReadExample use cases
Real sessions across science, finance, engineering, health & climate, and geopolitics & strategy — each with the original session, the outputs it produced, and a downloadable report for most.
ReadFAQ
Pricing, data security, file formats, collaboration, deployment options, and how to get started.
ReadSomething missing?
Documentation is thinner than the platform deserves and we know it. If you hit something that should be written down, tell us and we will write it.
support@k-dense.ai