Overview

Base URL

https://api.tcgvendorbase.com

All endpoints live under /api/v1/, accept GET only (except /cards/scan), and return JSON. Every request needs a bearer token from a registered app (see below). Responses are edge-cached, so repeat reads are fast.

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/cards?q=charizard&game=pokemon"

Authentication

Send your end user's identity token as a bearer token on every call. There is no shared API key — the X-API-Key header was retired and is no longer accepted.

Authorization: Bearer YOUR_ID_TOKEN

Each consuming app is registered by its token issuer. Both Supabase Auth and Firebase Auth are supported; the API verifies the token against that issuer's public keys, so no secret is shared and nothing sensitive ships in a browser bundle. Contact the API owner to register an app.

Registration also fixes the app's scopes. A token is valid but still refused if its app lacks the scope an endpoint requires:

Games & languages

Every product carries a game and a lang. Search endpoints accept both as optional filters (omit to search across all).

Currently available

Caching & prices

• Caching — GET responses are cached at the edge (s-maxage=3600, stale-while-revalidate=86400). Expect data up to ~1 hour old.
• Currency — *_usd fields come straight from TCGplayer (USD). market_myr is a snapshot computed at ingest using fx_rate (USD→MYR).
• Freshness — prices are refreshed from the source about daily; captured_at is the source's publish time.
• Variants — a product has one price row per sub_type_name (e.g. Holofoil, Normal, Reverse Holofoil, or Unopened for sealed).
• Prices — fetch with a batch /prices call using the product_ids returned by a search.

Try it — live console

Requests run against this same API (same origin, so no CORS setup needed). Your token is sent only to this API; tick remember to keep it in this browser only. Inputs are pre-filled with working examples.

—

Search cards

Fuzzy search over single cards, matching the card name, collector number and set name. 20 results per page, best match first.

Example

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/cards?q=charizard&game=pokemon&number=4/102"

{
  "page": 1,
  "total": 161,
  "results": [
    {
      "product_id": 84189,
      "group_id": 1863,
      "product_type": "card",
      "lang": "en",
      "game": "pokemon",
      "name": "Charizard",
      "card_number": "4/102",
      "image_url": "https://...",
      "tcgplayer_url": "https://...",
      "set_name": "SM Base Set",
      "similarity": 0.6
    }
  ]
}

Note on similarity: it measures the query against the card name only, while matching and ranking also consider the collector number and set name. A result can therefore be highly relevant yet carry a low similarity — q=crown zenith returns that set's cards with similarity near 0, since the query words are in the set name, not the card name. Use it to compare results within one response; don't threshold on it to decide what to show.

Get a card

Full detail for one card by its product_id. Returns 404 if the id is not a card (e.g. it's a sealed product).

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/cards/84189"

{
  "product_id": 84189,
  "group_id": 1863,
  "game": "pokemon",
  "lang": "en",
  "product_type": "card",
  "name": "Charizard",
  "card_number": "4/102",
  "image_url": "https://...",
  "tcgplayer_url": "https://...",
  "extended": [
    { "name": "Number", "displayName": "Number", "value": "4/102" },
    { "name": "Rarity", "displayName": "Rarity", "value": "Holo Rare" }
  ]
}

Search sealed

Same as Search cards but limited to sealed products (booster boxes, ETBs, bundles, …). Accepts q (required), game, lang, page.

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/sealed?q=booster%20box&game=pokemon"

Get a sealed product

Detail for one sealed product by product_id. Returns 404 if the id is not sealed. Same response shape as Get a card (with product_type: "sealed").

Search all products

Searches cards and sealed together; each result carries its own product_type. Use the optional type filter to narrow.

List sets

Every set (TCGplayer group), newest published first. Use a set's group_id to list its cards via Cards in a set.

Example

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/sets?lang=en"

{
  "results": [
    {
      "group_id": 1863,
      "name": "SM Base Set",
      "abbreviation": "SM",
      "lang": "en",
      "category_id": 3,
      "published_on": "2017-02-03"
    }
  ]
}

Cards in a set

Every card in a set by its group_id, paginated 60 per page and ordered by collector number. Same product shape as Search cards (without similarity).

Example

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/sets/1863/cards?page=1"

{
  "page": 1,
  "total": 102,
  "results": [
    {
      "product_id": 84189,
      "game": "pokemon",
      "lang": "en",
      "name": "Charizard",
      "card_number": "4/102",
      "image_url": "https://...",
      "image_url_fallback": "https://...",
      "tcgplayer_url": "https://...",
      "set_name": "SM Base Set"
    }
  ]
}

Batch prices

Current prices for up to 200 products in one call. Returns one row per variant; works for cards and sealed alike.

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/prices?ids=84189,91601"

{
  "results": [
    {
      "product_id": 84189,
      "sub_type_name": "Holofoil",
      "product_type": "card",
      "lang": "en",
      "game": "pokemon",
      "market_usd": 138.08,
      "market_myr": 547.86,
      "low_usd": 105.00,
      "mid_usd": 572.12,
      "high_usd": 1390.55,
      "direct_low_usd": 999.99,
      "fx_rate": 3.9677,
      "captured_at": "2026-05-30T20:04:57+00:00"
    }
  ]
}

Price history

One series per variant, oldest first. A row is appended only on the days the market price actually changed.

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/prices?ids=84189&history=1&days=90"

{
  "days": 90,
  "results": [
    { "product_id": 84189, "sub_type_name": "Holofoil", "d": "2026-05-22", "market_usd": 131.40, "market_myr": 533.52 },
    { "product_id": 84189, "sub_type_name": "Holofoil", "d": "2026-06-04", "market_usd": 138.08, "market_myr": 560.63 }
  ]
}

Reading the series. Three things will trip you up if you treat it as a dense daily feed:

• Gaps mean the price held steady. A row is only written when the market price actually moves, so consecutive dates are usually days or weeks apart. Draw it as a step line, carrying each value forward until the next point.
• The first point may predate your window. If a card's price last moved before days ago, that older point is still returned as the opening value — otherwise the chart would render empty for a card that simply hasn't moved. Its d can therefore be older than days.
• A null market_usd is meaningful. It marks the day the card stopped having a market price at all, which is different from holding steady.

market_myr is converted at that day's USD→MYR rate, not today's, so the series is comparable over time. Weekend and holiday dates use the last published weekday rate.

Chart series (spark=1)

Built for a card detail page: one request serves all five windows. Two series come back per printing, and their point counts nest, so switching window is a slice of what you already have — no second request.

curl -H "Authorization: Bearer YOUR_ID_TOKEN" \
  "https://api.tcgvendorbase.com/api/v1/prices?ids=274439&spark=1"

{
  "as_of": "2026-08-19",
  "windows": {
    "7d":  { "series": "daily",  "take": 7  },
    "30d": { "series": "daily",  "take": 30 },
    "3m":  { "series": "weekly", "take": 13 },
    "6m":  { "series": "weekly", "take": 26 },
    "1y":  { "series": "weekly", "take": 53 }
  },
  "results": [
    {
      "product_id": 274439,
      "sub_type_name": "Holofoil",
      "daily":  { "covered_since": "2026-07-21", "points": [ ... 30 points ... ] },
      "weekly": { "covered_since": "2026-01-07", "points": [
        { "d": "2026-08-19", "close_usd": 4.34, "low_usd": 3.64,
          "high_usd": 4.42, "close_myr": 17.61, "fx_rate": 4.0579 }
      ]}
    }
  ]
}

Rendering it. Take the last take points of the named series — e.g. 6m is the last 26 weekly points. Plot close_usd as the line.

• Use low_usd/high_usd, don't ignore them. They are the extremes actually reached inside that bucket. On weekly points the close can understate the week's high by 30–40%, so a line drawn from closes alone hides real spikes. Daily points hold one observation each, so their band is zero-width by construction.
• Series can be shorter than take. A printing that didn't exist a year ago has no year of history — about a fifth of the catalogue is younger than six months. covered_since is the first date present; label the chart from it (“Since Mar”) rather than claiming a full window.
• Points are evenly spaced, not one-per-change. Each is the price in force on that date, so the x-axis is uniform and safe to plot directly — unlike history=1, which is sparse and needs forward-filling.
• low_myr/high_myr aren't sent; multiply by that point's fx_rate.

Response fields

Product (search results & detail)

Price

Price history (history=1)

Errors

Errors return a JSON body {"error":"…"} with the matching status code.