Quickstart

# Today's gold price in INR, from the edge
curl -s https://snapdata.dev/api/v1/gold/in/latest.json | jq '.observations[]'

# A permanent, immutable URL, the same bytes in 2029
curl -s https://snapdata.dev/api/v1/gold/in/2026-08-11.json

# Everything, one request
curl -s https://snapdata.dev/api/v1/all/latest.json

# Token-cheap plain text
curl -s https://snapdata.dev/api/v1/gold/in/latest.txt

# Gold, in, 2026-08-11
Status: provisional | Trading day: yes | Calendar: IN-BULLION
Source: India Bullion and Jewellers Association (IBJA), via cmsapi.groww.in

XAU.18K  ₹11,460.20 /g
XAU.22K  ₹13,996.80 /g
XAU.24K  ₹15,280.30 /g

Prev (2026-08-10): XAU.24K ₹15,063.60  (+1.44%)

ETAG=$(curl -sI https://snapdata.dev/api/v1/all/latest.json | grep -i '^etag:' | cut -d' ' -f2- | tr -d '\r')
curl -s -o /dev/null -w '%{http_code}\n' -H "If-None-Match: $ETAG" \
  https://snapdata.dev/api/v1/all/latest.json
# 304

const res = await fetch("https://snapdata.dev/api/v1/all/latest.json");
const { observations, generated_at } = await res.json();

for (const o of observations) {
  // Never assume a value exists. A closed market emits an explicit row with
  // value: null and is_trading_day: false, that is the contract, not a bug.
  if (o.value === null) {
    console.log(`${o.instrument_id}, no value (${o.status})`);
    continue;
  }
  console.log(`${o.instrument_id}: ${o.value} [${o.status}]`);
}

const observed = observations.filter(
  (o) => o.value !== null && o.filled_from === null,
);

import urllib.request, json

url = "https://snapdata.dev/api/v1/crude/world/series/30d.json"
with urllib.request.urlopen(url) as r:
    data = json.load(r)

# Non-trading days are present as rows, so a DataFrame gets a real NaN on a
# holiday rather than a silently missing index entry.
rows = [
    (o["date"], o["value"])
    for o in data["observations"]
    if o["instrument_id"] == "BRENT.USD.BBL"
]

for date, value in rows[-7:]:
    print(date, value if value is not None else "closed")

import pandas as pd

df = pd.read_csv("https://snapdata.dev/api/v1/crude/world/series/30d.csv")
df = df[df.instrument_id == "BRENT.USD.BBL"]
df["date"] = pd.to_datetime(df["date"])

# `value` is genuinely NaN on closed days, not a forward-filled duplicate.
print(df.set_index("date")["value"].resample("W").mean())

https://snapdata.dev/llms.txt        the URL grammar, date rules, instruments
https://snapdata.dev/llms-full.txt   the same, plus current values and source health
https://snapdata.dev/openapi.json    templated paths, four shapes, not thousands of URLs

curl -s https://snapdata.dev/api/v1/all/latest.txt

curl -s https://snapdata.dev/api/v1/meta.json | jq '{status, stale_datasets}'

{
  "status": "degraded",
  "stale_datasets": ["equity-indices"]
}