mike_c just transformed a 1989 Ken Griffey Jr. rookie card 3 users watching your card category right now sarah_t upgraded to Pro — unlocked 4K + batch processing 47 cards transformed in the last hour · 89% quality gate pass rate ⏰ Pro plan at $9/mo — 12 spots left at launch price mike_c just transformed a 1989 Ken Griffey Jr. rookie card 3 users watching your card category right now sarah_t upgraded to Pro — unlocked 4K + batch processing 47 cards transformed in the last hour · 89% quality gate pass rate ⏰ Pro plan at $9/mo — 12 spots left at launch price
Image Capture & Upload
Drop sports card image here
JPG, PNG, WebP · Max 10MB · High-res recommended
Card Metadata
No fal.ai key — pipeline will simulate video.
Live Pipeline Monitor
Pipeline Ready
Upload a card and launch the transformation to see live stage progress.
Pipeline Architecture
Live system health + 7-stage AI pipeline breakdown
0
Total Assets
0
Completed
—
Avg Quality
—%
Success Rate
System HealthChecking...
Provider
fal.ai WAN 2.1
Model
fal-ai/wan-i2v
Last Checked
—
7-Stage Pipeline
① Upload & Ingest
Base64 image received, asset record created, ID assigned.
② Image Quality Analysis
Sharpness, brightness, contrast, noise scoring. Score <65 = rejected.
③ Quality Gate
Policy enforcement — bad images blocked before AI credit is spent.
④ Prompt Intelligence
Sport + player + era + condition → structured WAN 2.1 animation prompt.
⑤ AI Video Generation
fal.ai WAN 2.1 async job. 480p $0.20 · 720p $0.40 · ~60s delivery.
⑥ Output Validation
Optical flow + jitter + motion quality scoring on rendered video.
⑦ Media Registry
Full provenance chain: source → quality → prompt → job ID → video.
Tech Stack
RuntimeCloudflare Workers (edge)
FrameworkHono v4
AI Videofal.ai WAN 2.1
Model IDfal-ai/wan-i2v
BuildVite + TypeScript
DeployCloudflare Pages
StorageIn-memory (edge)
Queuefal.ai async queue + polling
480p cost$0.20 / video
720p cost$0.40 / video
Media Registry
Full provenance and lifecycle table for all assets
ID Name Sport Status Quality Video Duration Created