Evidence Gate
Front and back are mandatory. Low-resolution, cropped, blurred, or upscaled images are downgraded before any grade estimate.
CCA Matrix v1 ยท AI pre-grading before PSA
Know if a raw card is worth grading before you mail it.
CardCheck.ai turns front, back, corner, edge, and surface photos into a CCA AI Pre-Grading Report: PSA-style grade range, condition matrix, evidence limits, and a submit-or-skip decision.
GLM vision reads visible card evidence
CCA Matrix explains the grade range
QR record keeps a shareable report trail
CCA-20260701 sample
Estimated PSA range: 8-9
Back-edge whitening and centering risk make a PSA 10 outcome unlikely from the uploaded photos.
CardCheck.ai is built for one practical decision: should this raw card be submitted for grading, sold raw, kept, or skipped?
CCA report system
A structured report is harder to copy than a single AI score.
Each CCA report combines image evidence quality, card identity, condition signals, grade range, submit decision, QR access, and evidence fingerprints. The product is the decision trail, not only the model call.
GLM Vision Pass
The model reads visible identity, front/back consistency, typography, border, color, print, holo, and condition clues.
CCA Matrix
Centering, corners, edges, and surface are scored as separate dimensions with confidence and grade-cap notes.
Submit Decision
The report ends with strong candidate, maybe submit, skip, manual review, or recapture required.
Pre-grading flow
Designed to help collectors avoid bad submissions.
Capture
Front and back are mandatory. Corner close-ups and angled surface photos improve grade confidence.
Identify
The system reads name, set, number, language, year, and variant signals before estimating value impact.
Pre-grade
Centering, corners, edges, and visible surface issues are scored with conservative confidence boundaries.
Decide
The report recommends skip, maybe submit, or strong candidate based on grade range and grading economics.
CCA AI Pre-Grading Report
A report that explains the grade risk and the decision.
Every check produces a CCA record with card identity, photo quality, estimated grade range, condition matrix, submit recommendation, limitations, evidence fingerprints, and a QR link.
- PSA/BGS/CGC-style range, not a guaranteed official grade
- Centering, corners, edges, and surface condition matrix
- Submit, maybe, skip, manual-review, or recapture recommendation
- Photo limits and evidence fingerprints so the report has a trace
Trust model
Useful estimates without pretending to be PSA.
Evidence-first intake
The product rejects weak photos and asks for recapture instead of manufacturing a confident grade from weak inputs.
Clear boundary
CCA reports are AI-assisted estimates, not official PSA, BGS, CGC, SGC, or TAG results.
Data moat
Collectors can upload final grading results later, helping calibrate predicted ranges against real PSA/BGS/CGC outcomes.
Launch offer
Start with a photo gate, then pay for decisions.
Free checks tell collectors whether photos are usable. Paid reports add grade range, condition explanation, and submit-or-skip guidance.
Photo quality check, basic identity, and recapture guidance.
Estimated grade range, condition breakdown, and submit recommendation.
Monthly credits, report library, and ROI tools for frequent submissions.
Batch pre-screening, CSV export, and team workflow for sellers.