Loading apps…
Workflow
Make it in Vellm, then take mp4 video into LocateAnything. Matched from what each app's own site says it takes in and gives out.
Step 2
Takes mp4 video from Vellm. Find objects and text in images using natural language queries.