Loading apps…
Workflow
Make it in Archsense, then take diagram into SmolVLM-256M. Matched from what each app's own site says it takes in and gives out.
Step 2
Takes diagram from Archsense. Upload images and ask questions about their content using a small vision-language model.