Loading...
Loading...
CGI Vision AI scans 47 inspection points across centering, corners, edges, and surface. Catch centering and edge issues before tying up cards for months.
Both grades run on CGI Vision AI and cost 1 credit. The difference is what you hand the AI to work with.
One full-card photo. CGI Vision AI reads the entire card surface and returns PSA, BGS & CGC predictions with a per-pillar breakdown in about 60 seconds. The standard for everyday pre-screening.
Eight macro close-ups — one per corner, front and back — instead of one full-card photo. The extra detail gives CGI Vision AI sharper whitening, edge-wear, and fraying detection — useful when the difference between a 9 and a 10 is in the corners.
Snap or upload a photo. On-screen alignment guides flag lighting and angle issues before you submit — so the AI scores what a grader would actually see. Intelligence Grade uses one full-card photo. Forensic Capture (Pro & Dealer) uses eight macro close-ups — one per corner, front and back — for sharper whitening, edge-wear, and fraying detection. Both cost 1 credit.
Guided capture system
Alignment guides flag lighting and angle
CGI Vision AI maps every pillar across 16 zones — the same condition breakdown professional graders use, turned into measurable data points you can act on.
Powered by AI trained on thousands of card images
| Feature | CardGrade.io | Traditional |
|---|---|---|
| Turnaround time | ~60 seconds | 30–60 days |
| Cost per card | $0.42–$1.00 | $20–$50+ |
| Condition breakdown | 47 inspection points | Final grade only |
| Market value estimate | Included | Not included |
| Grade companies covered | PSA, BGS & CGC | One company per submission |
| Guarantee | 2-grade credit-back | None |
CardGrade.io is instant pre-screening — know which cards are worth submitting before you wait months.
Deep learning models trained on thousands of card images recognize the surface and structural patterns that determine grade outcomes.
Pixel-level analysis identifies card boundaries, borders, and defects — reading the same physical attributes a professional grader examines.
Our models improve over time as we incorporate new graded card data and community feedback.