Amazon Price History Chart API: graphs=true for Instant PNG Charts

Adding graphs=true to an /amazon-product-statistics call generates a watermarked PNG chart URL for every history series that has data, for the exact same single credit as the plain call - no second request, no extra charge. The URLs land in data.graphs, keyed by series name, and a series with nothing to chart is simply absent from that dict rather than null or broken.

Which series can actually get a chart?

How do I pull one chart URL, or every chart a call actually produced?

chart_url() reads one named series and returns None when that series has no chart, instead of raising. all_available_charts() returns whatever data.graphs actually contains, which can be an empty dict for a product with no trackable history at all. get() and new_session() come from the Python quickstart; the raw series themselves are covered in the Price History API guide.

chart_url.py - reading graphs=true chart URLs

from sellermagnet import get, new_session


def chart_url(session, asin: str, marketplace_id: str, series: str) -> str | None:
    """A watermarked PNG chart URL for one history series, for the SAME single
    credit as a plain statistics call (graphs=true adds it, no extra charge).
    Returns None when that series has no data to chart - not every series has
    a chart on every ASIN."""
    data = get(session, "amazon-product-statistics", asin=asin, marketplaceId=marketplace_id, graphs="true")
    return data.get("graphs", {}).get(series)


def all_available_charts(session, asin: str, marketplace_id: str) -> dict:
    """Every chart URL this call actually produced, keyed by series name.
    Series with no data (e.g. a product Amazon never resold) are simply
    absent from the dict, not present with a null or empty-string URL."""
    data = get(session, "amazon-product-statistics", asin=asin, marketplaceId=marketplace_id, graphs="true")
    return data.get("graphs", {})


if __name__ == "__main__":
    session = new_session()
    print(chart_url(session, "B0CL61F39H", "ATVPDKIKX0DER", "salesRankHistory"))
    print(chart_url(session, "B0CL61F39H", "ATVPDKIKX0DER", "amazonAsSellerPriceHistory"))
    print(all_available_charts(session, "B0NOCHARTS0", "ATVPDKIKX0DER"))

Example output (illustrative ASIN)

https://s3.example/charts/B0CL61F39H_sales_rank.png
None
{}

Trimmed data.graphs (example values)

{
  "graphs": {
    "salesRankHistory": "https://.../B0CL61F39H_sales_rank.png",
    "buyBoxPriceHistory": "https://.../B0CL61F39H_buybox.png"
  }
}

Is a chart URL ever a reason to skip the raw series?

No - graphs=true is a convenience, not a replacement. The raw arrays stay in the response exactly as without the flag, so you can keep reading salesRankHistory or the price series programmatically for your own logic (alerts, trend math) while using the chart only for a human-facing view. The BSR by Category guide and the pandas guide both work from the same raw series this endpoint returns.

Bottom line: add graphs=true for an instant watermarked chart at no extra cost, read data.graphs defensively since a series can be absent, and fall back to plotting the raw series yourself whenever you need an unwatermarked or custom chart.