Your project data never leaves your machine
Element names, file names, level identifiers, and discipline tags stay on your machine. The AI model that generates the title runs as a separate process on your computer, and no part of your clash data is sent to ClashWise or any third party.
No quotas, and nothing metering you
Update thousands of clashes in one batch, re-run a whole project, or iterate on a naming standard until it is right. Nothing counts your titles, so there is no cap to run into and no bill that scales with how much you use it.
Works once the AI is downloaded
After the one-time AI download, no internet is required for clash titling. Run on a site with restricted internet, in a regional BIM hub with patchy WAN, or anywhere a corporate firewall blocks AI services.
Ten curated open-source AI models
The catalog spans Google, Alibaba, Meta, Microsoft, Mistral AI, and OpenAI, from sizes small enough for a laptop up to workstation-class options, so you can fit the model to the machine you have and switch whenever you want.
Frequently asked questions
Is local AI as good as cloud AI?
For the vast majority of AEC clash titles, yes. Cloud uses a larger frontier AI so it has a slight quality edge on unusually ambiguous clashes. Local is consistently strong on the typical case, especially with naming standards.
Does any data leave my machine?
With local mode and no cloud fallback enabled, no clash data leaves your machine. Only usage analytics such as the count of titles generated flow back to ClashWise. They never include element names, file paths, or clash content.
Will this slow down Navisworks?
No. The AI model runs as a separate process, so Navisworks keeps its own memory and responsiveness. If the model ever crashes, the plug-in just relaunches it.