Oil price API: WTI, Brent and natural gas prices in JSON

An oil price API returns crude and gas prices as JSON your code can read. With TickerLayer, GET /commodities/quote/WTIUSD returns the WTI crude bid, ask and a Unix-millisecond timestamp; swap the symbol for BRENTUSD to get Brent or NGASUSD for natural gas. The commodities.quotes WebSocket channel streams the same fields, and the aggregates route returns daily and intraday bars.

Energy sits in the commodities API next to metals and grains, under one key and one set of routes. What makes oil harder than a stock quote is everything around the number: two competing benchmarks (compared in WTI vs Brent), prices set in futures markets, contracts that expire every month, and units that change from product to product. This guide takes each in turn, with code that ran against the live API on 28 September 2026.

Energy symbols and their units

The split between the last two columns is deliberate. TickerLayer labels a unit only where the public contract documents one, and the energy references are documented as dollar prices. The magnitudes match the conventions (WTI near $96 and natural gas near $3.10 on 28 September 2026), but pin the unit in your own code as a named constant instead of letting a bare float travel through your system.

Your first crude oil price API request

curl -sS "https://api.tickerlayer.com/commodities/quote/WTIUSD" \
  -H "x-api-key: $TICKERLAYER_API_KEY"

{
  "symbol": "WTIUSD",1
  "bid": 96.194,2
  "ask": 96.242,3
  "bid_size": 40,4
  "ask_size": 44,
  "timestamp": 17905911284595
}

For several benchmarks at once, loop over the symbols. This script prints each quote with its own observation time, then the Brent minus WTI spread. It needs Python 3.9 or newer and pip install requests.

import os
from datetime import datetime, timezone

import requests

API = "https://api.tickerlayer.com"
HEADERS = {"x-api-key": os.environ["TICKERLAYER_API_KEY"]}
ENERGY = {"WTIUSD": "WTI crude", "BRENTUSD": "Brent crude", "NGASUSD": "Natural gas"}


def quote(symbol: str) -> dict:
    r = requests.get(f"{API}/commodities/quote/{symbol}", headers=HEADERS, timeout=10)
    if r.status_code == 403:
        raise SystemExit("This key has no commodities access: see tickerlayer.com/pricing")
    r.raise_for_status()
    return r.json()


mids = {}
for symbol, name in ENERGY.items():
    q = quote(symbol)
    bid, ask = float(q["bid"]), float(q["ask"])
    mids[symbol] = (bid + ask) / 2
    at = datetime.fromtimestamp(q["timestamp"] / 1000, tz=timezone.utc)
    print(f"{name:<12} {symbol:<9} bid {bid:>8.3f}  ask {ask:>8.3f}  {at:%H:%M:%S} UTC")

print(f"Brent minus WTI: {mids['BRENTUSD'] - mids['WTIUSD']:.2f} USD")

WTI crude    WTIUSD    bid   96.146  ask   96.194  10:56:27 UTC
Brent crude  BRENTUSD  bid  108.360  ask  108.410  10:56:22 UTC
Natural gas  NGASUSD   bid    3.101  ask    3.103  10:56:14 UTC
Brent minus WTI: 12.21 USD

Look at the three timestamps. Each quote carries its own observation time, and the natural gas quote here was 13 seconds older than the WTI one. Before you subtract two prices, check that they belong to roughly the same moment; for a spread that feeds a decision, compare the timestamp values and skip the calculation when they drift too far apart.

The front month: what one oil price really is

Almost every headline oil price is a futures price. Crude changes hands in physical cargoes, but the number the world watches is set by the nearest futures contract, the front month. Each contract has a delivery month and a last trading day, and when the front contract expires, the next one takes its place.

A single daily series, the kind a chart or a backtest wants, has to cross from one contract to the next. The commodities data-quality page describes TickerLayer’s energy symbols as continuous references and treats spot, front-month and continuous futures as different instruments. That is the right mental model: WTIUSD is one unbroken series, not a single dated contract.

The size of the step equals the price gap between the two contracts, so it depends on the curve. In a calm, gently sloping market it is a few cents. In a supply squeeze, when near-term barrels are scarce and later ones are cheaper, it can be dollars. The contango and backwardation explainer shows the curve shapes behind it.

Units: barrels, MMBtu and cents

Energy prices come in different quantities, and mixing them is how a dashboard ends up plotting natural gas and crude on one axis. These are the conversions that come up in practice:

Natural gas is the most regional of the three. US gas is conventionally quoted in dollars per MMBtu, while European gas trades in euros per megawatt-hour, and one megawatt-hour is about 3.412 MMBtu. The same gas priced per megawatt-hour therefore shows a number about 3.4 times larger than per MMBtu, before any currency or real price difference. Convert both sides to one unit and one currency before you compare regions.

Softs and grains add a second trap, because their convention is cents. SUGARUSD is documented in US cents per pound, so a value of 18.542 means $0.18542 a pound, and CORNUSD, WHEATUSD and SOYBEANUSD are in US cents per bushel. Plot sugar next to a dollar-quoted series without dividing by 100 and it is off by a factor of a hundred. The troy ounce and commodity units explainer has the full table, metals included.

Historical oil price data

Daily and intraday bars come from the aggregates route, under the same symbol as the live quote. Bar sizes are 1, 5 and 15 minutes, 1 and 4 hours, and 1 day; dates are UTC YYYY-MM-DD; limit goes up to 5,000, and next_offset pages through longer ranges. History depth is two years on Individual plans and ten on Business.

curl -sS "https://api.tickerlayer.com/commodities/agg/WTIUSD/1/day/2026-09-13/2026-09-25?sort=asc" \
  -H "x-api-key: $TICKERLAYER_API_KEY"

The chart is a real, volatile month: WTI swung between $88.72 and $106.72 in two weeks. It also shows why the contract calendar belongs on the chart. The slide into 22 September, the last trading day of the October contract, can blend a falling market with the switch to November, and only the curve on that day can tell you how much was which.

Streaming oil prices over WebSocket

For a ticker or an alert, subscribe to commodities.quotes on the stream instead of polling. Right after the subscribe acknowledgement you get the last known value per symbol, marked "snapshot": true; after that, frames arrive as the reference moves. WebSocket access comes with paid plans.

{
  "action": "subscribe",
  "channels": ["commodities.quotes"],
  "symbols": ["WTIUSD", "BRENTUSD", "NGASUSD"]
}

{
  "type": "quote",1
  "channel": "commodities.quotes",
  "asset": "commodities",
  "symbol": "WTIUSD",
  "bid": 96.163,2
  "ask": 96.211,
  "bid_size": 51,
  "ask_size": 97,
  "ts": 1790593058637,3
  "timestamp": 1790593058637,
  "snapshot": true4
}

The gold price API guide has a complete Node.js streaming script that works unchanged for oil once you swap the symbols, and the WebSocket docs cover connection limits, error frames and reconnects.

Polling budget: how many requests an oil dashboard needs

REST is the simplest way to start, and the quota math decides how long it stays simple. Three energy symbols refreshed every 10 seconds for 23 hours a day is 24,840 requests a day, about 546,000 over 22 trading days. That is more than twice the 250,000 monthly calls on an Individual plan, for one small panel.

Two fixes keep the budget sane. Poll the snapshot only while someone is looking at the page, and move always-on panels and alerts to one WebSocket connection, which costs no REST calls at all. Every authenticated REST response also carries X-RateLimit-Limit and X-RateLimit-Remaining, so read your per-second ceiling from the headers instead of hard-coding it, and back off on a 429 for as long as Retry-After says.

What to look for in oil price data

On the provider side, a usable commodity price API documents its symbols and units, returns bid and ask, timestamps every quote and serves history under the same symbol. TickerLayer’s energy data is indicative reference pricing for applications and analytics, not an exchange settlement feed, and the market data disclaimer spells out what that means.