Fetching Intraday Session Forecasts from Nexenergie

After the day-ahead market closes, OMIE runs a series of intraday adjustment sessions that allow market participants to revise their positions as real-time conditions evolve. Nexenergie generates price forecasts for intraday sessions 1, 2, and 3, each covering a progressively shorter window of remaining delivery hours. This guide explains how to call the intraday endpoint, interpret the session-scoped response, and decide which session to query for your use case.

Session Numbering

Nexenergie follows the OMIE session numbering convention. Each session opens and closes at a fixed time and covers only the delivery hours that remain after its gate closure.

Session 1 — opens early morning and covers delivery hours from mid-morning onward. It is the first intraday opportunity to adjust positions after the day-ahead market settlement.
Session 2 — opens late morning and covers delivery hours from early afternoon onward. Fewer hours remain, and the model has access to fresher real-time fundamentals.
Session 3 — opens early afternoon and covers delivery hours from early evening through midnight. This is typically the session with the smallest coverage window but also the one that benefits from the most up-to-date ENTSO-E and weather data.

Calling the Intraday Endpoint

To fetch intraday forecasts, call GET /forecasts/intraday/{session}, replacing {session} with 1, 2, or 3. Pass the delivery date as the required date query parameter in YYYY-MM-DD format.

curl --request GET \
  --url "https://app.nexenergie.ai/api/v1/forecasts/intraday/2?date=2024-11-15" \
  --header "Authorization: Bearer <your_token>"

import requests

BASE_URL = "https://app.nexenergie.ai/api/v1"
TOKEN = "<your_token>"
SESSION = 2

response = requests.get(
    f"{BASE_URL}/forecasts/intraday/{SESSION}",
    params={"date": "2024-11-15"},
    headers={"Authorization": f"Bearer {TOKEN}"},
)
response.raise_for_status()
forecast = response.json()

print(f"Session {forecast['session']} — {len(forecast['forecasts'])} hours covered")
for entry in forecast["forecasts"]:
    print(f"  Hour {entry['hour']:02d}:00 → {entry['forecast_price']} EUR/MWh")

Understanding the Response

A successful 200 OK response mirrors the day-ahead structure but includes a session field and only contains entries for the hours covered by that session. The example below shows a session 2 forecast for an afternoon-to-midnight window.

Example Response (Session 2)

{
  "market": "intraday",
  "session": 2,
  "date": "2024-11-15",
  "model_version": "v2.4.1",
  "data_as_of": "2024-11-15T10:15:00Z",
  "forecasts": [
    { "hour": 12, "forecast_price": 81.30, "lower_bound": 73.5, "upper_bound": 89.1, "unit": "EUR/MWh" },
    { "hour": 13, "forecast_price": 85.74, "lower_bound": 77.2, "upper_bound": 94.3, "unit": "EUR/MWh" },
    { "hour": 14, "forecast_price": 90.12, "lower_bound": 81.8, "upper_bound": 98.5, "unit": "EUR/MWh" },
    { "hour": 15, "forecast_price": 88.45, "lower_bound": 80.1, "upper_bound": 96.8, "unit": "EUR/MWh" },
    { "hour": 16, "forecast_price": 86.20, "lower_bound": 78.4, "upper_bound": 94.0, "unit": "EUR/MWh" },
    { "hour": 17, "forecast_price": 83.67, "lower_bound": 75.9, "upper_bound": 91.4, "unit": "EUR/MWh" },
    { "hour": 18, "forecast_price": 79.55, "lower_bound": 71.3, "upper_bound": 87.8, "unit": "EUR/MWh" },
    { "hour": 19, "forecast_price": 75.88, "lower_bound": 68.0, "upper_bound": 83.7, "unit": "EUR/MWh" },
    { "hour": 20, "forecast_price": 72.40, "lower_bound": 64.6, "upper_bound": 80.2, "unit": "EUR/MWh" },
    { "hour": 21, "forecast_price": 68.95, "lower_bound": 61.2, "upper_bound": 76.7, "unit": "EUR/MWh" },
    { "hour": 22, "forecast_price": 64.30, "lower_bound": 57.0, "upper_bound": 71.6, "unit": "EUR/MWh" },
    { "hour": 23, "forecast_price": 60.10, "lower_bound": 53.4, "upper_bound": 66.8, "unit": "EUR/MWh" }
  ]
}

The session field confirms which intraday session the forecast belongs to. The data_as_of timestamp reflects the latest real-time inputs available at model run time, which will be more recent than the equivalent day-ahead timestamp for the same delivery date.

Choosing the Right Session

Use the table below to select the session that matches your trading workflow.

Session	Typical Hour Coverage	When to Use
1	Hours 8–23 (approx.)	Early-morning position adjustments after day-ahead settlement; first intraday look at the full remaining day.
2	Hours 12–23 (approx.)	Late-morning refinement when solar and wind actualization data is available; recalibrating afternoon positions.
3	Hours 16–23 (approx.)	Near-real-time fine-tuning for evening hours; highest data freshness but narrowest coverage window.

Session 3 forecasts are only generated once the model has ingested sufficient updated ENTSO-E and weather telemetry, which typically occurs in the early afternoon. If you query session 3 before it has been generated for the requested date, the API returns a 404 Not Found response. Build a retry loop with back-off into any automation that depends on session 3 data.

If you need a full 24-hour price view to plan your daily strategy, use the day-ahead endpoint instead. Reserve intraday session calls for workflows that specifically require near-term position adjustments or the incremental accuracy gains that later sessions provide.

Get Started

Core Concepts

Guides

Account & Settings

Fetching Intraday Session Forecasts from Nexenergie

Session Numbering

Calling the Intraday Endpoint

Understanding the Response

Choosing the Right Session

​Session Numbering

​Calling the Intraday Endpoint

​Understanding the Response

​Choosing the Right Session

Session Numbering

Calling the Intraday Endpoint

Understanding the Response

Choosing the Right Session