AI Pokémon card grading and pre-grading

Get your TCGAI Score before you submit. From vintage WOTC Base Set to modern Scarlet & Violet, the AI measures centering, corners, edges and surface so you only send your top-score candidates.

2 free scans a month · No card required · TCGAI Score is an estimate, not an official grade

TCGAI.PRO 9.5
Stormlynx
#014/120 Submit
Stormlynx HP 120 STORM BEAST · STAGE 1 Static Pounce 30 Flip a coin. If heads, the target is stunned. Thunder Veil 110 This beast takes 20 less damage next turn. It hunts only when the sky starts to hum. TCGAI ORIGINALS · 014/120
Sample Card from our example set. Estimate, not an official grade.

Supported Pokémon sets and eras

The same four sub-grades, checked on Pokémon cards from every era. Standard and Japanese-size cards, raw or already in a slab.

  • Vintage WOTC Base Set, Jungle, Fossil, Team Rocket, Gym, Neo, e-Series
  • EX / GX era Ruby & Sapphire, Sun & Moon GX, Tag Team
  • V / VMAX / VSTAR Sword & Shield, Evolving Skies, Brilliant Stars
  • Scarlet & Violet 151, Obsidian Flames, Illustration Rares
  • Japanese cards Promos, VSTAR Universe, Shiny Treasure ex
  • Trophy & promo Illustrator, Trophy, Black Star Promos
  • Modern hits Alt Art, Special Art Rare, Gold, Full Art
  • All rarities Common to Secret Rare, any Pokémon card

What 2,550 Pokémon cards scored

Every Pokémon card graded on TCGAI.PRO from December 2025 to October 2026 (11 cards with no year are left out of the table). The age of the card matters more than anything else.

Card yearCardsAverageScored 10
1999 to 20023256.425.2%
2003 to 20161926.817.8%
2017 to 20224518.1812.2%
2023 to 20261,5718.7921.5%

TCGAI Scores are AI estimates, not official grades. Full figures and method: Pokémon grading data 2026.

AI pre-grading vs. sending it blind

Pre-screen your Pokémon cards before you pay a third-party grading fee.

Criterion Submit blind TCGAI.PRO
Time to a result 45 to 90 days Under 2 minutes
Cost per card $30 to $150+ Free tier, then from $6.99/mo
Consistency Depends on the grader Same rules on every card
Screening before you submit No, you pay either way Yes, filter out the 8s and 9s
Where the card stays In their vault for months In your hands
Holo scratches Checked by eye Micro-scratch scan

What the AI checks on Pokémon cards

What matters on Pokémon cards: holographic surfaces, texture rares and vintage WOTC printing patterns.

  1. Centering L/R · T/B

    Front and back ratios (55/45 for a 10). Off-centre past 65/35 caps the card at a 9 before anyone looks at the corners.

  2. Corners 8 tips

    Sharpness, whitening, dings. One soft tip is usually the whole difference between a 9 and a 10.

  3. Edges 8 edges

    Edge wear, chipping, roughness. Chipping hides in the border. It is the most common reason a clean card comes back a 9.

  4. Surface 2 faces

    Holo scratches, print lines, silvering. Print lines and holo scratches are invisible in a phone photo and fatal to gem mint.

Centering on Pokémon cards

Measures the exact pixel distance between borders and checks it against the tables graders use: 55/45 on the front and 75/25 on the back for a 10 (a 60/40 front gives a 9). Critical for WOTC-era cards with inconsistent cutting.

Holo scratch detection

Finds micro-scratches, holo-bleeding and silvering that are easy to miss by eye. Essential for Charizard VMAX, Illustration Rares and vintage 1st Edition holos.

Edge whitening analysis

Detects corner whitening and soft corners, and flags the minor chipping that would drop a Pokémon card from a TCGAI Score 10 to a 9.

Is your Pokémon card a 10? Find out free in about two minutes.

Grade a card free

Pokémon grading FAQ

The TCGAI Score is an AI estimate, not an official grade. We are not affiliated with any grading company.

Grade a card free
How accurate is the AI for Pokémon cards?

Our AI checks the same observable criteria professional graders look at (centering, corners, edges and surface) and produces a TCGAI Score with sub-grade detail. The TCGAI Score is an AI estimate, not an official grade. Third-party grades may differ. We are not affiliated with any third-party grading company.

Does it work on vintage WOTC Pokémon cards?

Yes. Vintage cards from Base Set, Jungle, Fossil, Team Rocket, Neo Genesis and the e-Series are checked on the same four sub-grades. WOTC-era cards often have inconsistent centering, which is measured from the borders like any other card. In our data, Pokémon cards from 1999 to 2002 averaged 6.42.

Can it detect holo scratches on Pokémon cards?

Yes. The surface analysis targets holographic surfaces: micro-scratches, holo-bleeding and silvering on foil and texture rare cards that are easy to miss by eye.

Should I send my Pokémon card for grading if the TCGAI Score is 9?

It depends on the card. A 9 can still be very valuable for high-demand cards like a 1st Edition Charizard, and the market estimate shows the price difference between grades. Remember that the TCGAI Score is an estimate; the grade you receive may differ.

Can I AI grade Japanese Pokémon cards?

Yes. Japanese Pokémon cards are supported, including VSTAR Universe, Shiny Treasure ex, Art Rares and vintage promos, on the same four sub-grades. In our data, Japanese Pokémon cards averaged 8.54 against 8.09 for English ones.

Which Pokémon sets rarely score a 10?

Among Pokémon sets with at least 30 cards graded on TCGAI.PRO (December 2025 to October 2026), the lowest shares of perfect 10s were Base Set (0%, average 6.32), Perfect Order (2.7%), Ascended Heroes (4.9%), Jungle and Celebrations (6.7% each) and Crown Zenith (8.8%, even with a 9.09 average). Modern 151 reached 31%. Pre-grade cards from these sets before paying for a submission.

How do I photograph holographic Pokémon cards?

Remove the sleeve, use a dark background and diffused, even lighting, and tilt the card slightly to minimize holo glare.