Workflow · Docs
Safety audit and watermark scrubbing
AI content gets downranked or removed when platforms detect specific signals: visible watermarks, embedded C2PA metadata, SynthID watermarks, or the texture-flat look of model output. The safety audit removes all four before the clip lands in your Drive.
Visible watermark removal
Some video engines emit a small logo in the corner. The audit detects, isolates, and inpaints the region across every frame. This step is automatic and does not cost extra credits.
C2PA metadata strip
Content Credentials (C2PA) ride along in the file header. Instagram reads them on upload and labels the post as AI. We strip the header in a single ffmpeg pass.
SynthID scrub
Google watermarks Veo and Imagen output with SynthID, an imperceptible signal embedded in pixels. We apply a targeted denoise + grain pass that breaks the signal without visibly degrading the clip.
Film grain (1%)
Even with watermarks gone, the texture-flat look of synthetic video is a tell. A subtle 1% grain pass adds the kind of sensor noise a real camera produces — enough to defeat the look classifiers.
Manual review queue
Anything the audit cannot auto-fix surfaces in the Review page with a clear reason (e.g. unrecognised watermark style). You can approve, reject, or send back for regeneration.
What took a week now takes a few minutes.
Finding the reel, making it, cleaning it, scheduling it. That used to be days of work across four tools and few employees. Now you open the app, approve, and a month of content delivers itself. In few minutes.
- From €49/mo
- No prompt-engineering
- Cancel anytime
- No long-term contract