croncopia/commodity-prices

Aggregated commodity prices, rebuilt every 30 minutes from independent providers and published as static JSON — one file per commodity, no API key required, no rate limits.

How to access the data?

Access is as simple as fetching the specific JSON file. There are multiple ways to do this depending on your caching and freshness needs — to target a specific commodity, change metals/gold to the relevant commodity path:

GitHub Pages — served as a static site, good if you want a less verbose url and a stable endpoint with predictable caching via GitHub's own CDN:

http://commodity.croncopia.com/latest/metals/gold.json

jsDelivr — CDN-cached globally, fast and reliable for production use (cache typically refreshes every ~12-24h, or instantly if you pin a commit SHA instead of a branch):

https://cdn.jsdelivr.net/gh/croncopia/commodity-prices/latest/metals/gold.json

Statically — alternative CDN proxy, similar caching behaviour to jsDelivr, useful as a fallback or for load distribution:

https://cdn.statically.io/gh/croncopia/commodity-prices/main/latest/metals/gold.json

Githack — serves files with correct Content-Type headers, handy if you're fetching this client-side and need proper MIME types rather than text/plain:

https://raw.githack.com/croncopia/commodity-prices/main/latest/metals/gold.json

GitHub raw — direct from the repo, always reflects the latest commit, no caching layer (subject to GitHub's rate limits on heavy use):

https://raw.githubusercontent.com/croncopia/commodity-prices/refs/heads/main/latest/metals/gold.json

Note

If you need guaranteed up-to-the-minute data, use the raw.githubusercontent.com link. If you're optimizing for speed/uptime and can tolerate slightly stale data, the CDN options (jsDelivr, Statically, Githack) are the better choice.

Data Structure

Each file holds a single commodity's USD price converted across the standard weight units, stamped with the moment it was generated:

{
  "base": "USD",
  "price": {
    "troy_ounce": 4264.570915110066,
    "gram": 137.10913871564566,
    "kilogram": 137109.13871564565,
    "ounce": 3886.9786952403006,
    "pound": 62191.659178688475,
    "metric_ton": 137109138.71564567
  },
  "sources": 3,
  "timestamp": "2026-06-18T16:28:30.666Z"
}

How it works?

1. Fetch — each source module pulls the latest price from its provider and normalises it to a common unit: USD per troy ounce.
2. Filter — only commodities on a defined whitelist pass into the aggregate, a single choke point that keeps unexpected or malformed symbols from a source out of the dataset.
3. Aggregate — sources are fetched concurrently and each may fail independently; the run continues as long as at least one source delivers. Per commodity, the prices are averaged — and when 3+ sources report, any price deviating more than 10% from the median is discarded as an outlier before averaging.
4. Distribute — one JSON file is written per commodity, with the averaged USD/troy-ounce price converted into the other weight units shown above.