Access to the Nodal Price Forecasts API
The Nodal Price Forecasts are available exclusively through our API URL here:
Access requires a valid API key and subscription to the appropriate service package.
1. What are the different endpoints I have access to with a subscription?
GET /prediction
Description: Nodal price forecasts for a specific node returning the predicted LMP for the selected node, the MCC, MEC and prediction distribution (deciles) of MCC and LMP values. This endpoint also returns LMP, MCC predictions and confidence factors for load zones and hubs.
Time zone: CST
Required Inputs:
Node Name - a single ERCOT node name for which the forecast is requested.
Example:
node=aeecDate (one of the following):
Operation Date - The date for which the forecast is generated (in Central Time). This will return 48 hours of data.
Example:
operationDate=2025-08-26Publication Date Timestamp - The timestamp indicates when the forecast was published (in Central Time). This will return 48 hours of data.
Example:
publicationDateTime=2025-08-26T09:00:00-05:00
GET /bulkprediction
Description: NPFs for all, or multiple nodes in one request returning predictions and confidence factors, LMP prediction distribution is not included. This endpoint also returns LMP, MCC predictions and confidence factors for load zones and hubs.
Time zone: CST
Required Inputs:
Node Name - one or more ERCOT node names for which the forecast is being
requested. If this is omitted all nodes will be returned.
Example:
nodes=amistad_all&nodes=aeecDate (one of the following):
Operation Date - The date for which the forecast is generated (in Central Time). This will return 48 hours of data.
Example:
operationDate=2025-08-26Publication Date Timestamp - The timestamp indicates when the forecast was published (in Central Time). This will return 48 hours of data.
Example:
publicationDateTime=2025-08-26T09:00:00-05:00
GET /context
Description: Nodal price forecasts for a specific node returning the predicted LMP for the selected node, the MCC, MEC and prediction distribution (deciles) of MCC and LMP values.
Time zone: CST
Required Inputs:
Node Name - a single ERCOT node name for which the forecast is requested.
Example:
node=aeecDate (one of the following):
Operation Date - The date for which the forecast is generated (in Central Time). This will return 48 hours of data.
Example:
operationDate=2025-08-26Publication Date Timestamp - The timestamp indicates when the forecast was published (in Central Time). This will return 48 hours of data.
Example:
publicationDateTime=2025-08-26T09:00:00-05:00
GET /constraintprediction
Description: Provides predicted constraints and shadow prices used in the calculation of the nodal prices.
Required Inputs:
Date (one of the following):
Operation Date - The date for which the forecast is generated (in Central Time). This will return 48 hours of data.
Example:
operationDate=2025-08-26Publication Date Timestamp - The timestamp indicates when the forecast was published (in Central Time). This will return 48 hours of data.
Example:
publicationDateTime=2025-08-26T09:00:00-05:00
GET /nodedetails
Description: List of nodes where predictions have been generated, the timestamp where predictions last generated for each node, the start and end date range of predictions on the last run for each node.
Required Inputs:
Node Names - one or more ERCOT node names for which the forecast is being
requested. If this is omitted all nodes will be returned.
Example:
nodes=amistad_all&nodes=aeec
2. Technical Reference
2.1 What data format does the API return?
The API returns the data in JSON format.
2.2 What time zone are the predictions provided in?
All predictions are provided in Central Prevailing Time
2.3 What authentication method does the API use?
The API uses API key authentication. You must include your API key in the request header for every call. For example:
GET /example-endpoint
Host: api.example.com
Authorization: Bearer YOUR_API_KEY
Replace YOUR_API_KEY with the key provided in your account settings. Requests without a valid API key will return an error.
2.4 What are common API errors I might encounter?
Common errors typically relate to:
Incorrect input parameters
Timeouts depending on the tools used
Authentication issues
Sending specific error codes to our technical team through support@woodmac.com will help us quickly troubleshoot your error.
2.5 How often do you variables update?
3. Data Analysis & Tools
3.1 Can I pull the data into Excel?
Yes, you can pull data into Excel using the API. However, we recommend using more robust solutions for large datasets due to the substantial amount of data available. We have created python toolkits to help you spin up quick insights and visualizations for the data.
3.2 Do you provide any tools or toolkits to help with data visualization?
Yes, we have created Python toolkits specifically designed for our customers to create quick visualizations and rapidly prototype solutions. These toolkits allow you to get up and running quickly with the data without having to build everything from scratch. For access to these toolkits and documentation on how to use them, please contact support@woodmac.com.
3.3 How accurate is your forecast?
We are continually enhancing the model to improve our methodology and accuracy. Building a retrospective data set will allow for accuracy scoring in the future.
3.4 How can I compare actuals to forecasts?
We have created a Python toolkit specifically for comparing LMP forecasts to actual values. Please contact support (support@woodmac.com) for access to this toolkit or contact your account manager or CSM who will be able to forward this on.
3.5 Can I set up automated data pulls/scheduling?
Yes. The API is designed for programmatic access, so you can build automated jobs to fetch data on a schedule. Common patterns include:
Client libraries / scripts – Write scripts in Python, JavaScript, or another language to call the API and handle responses.
Cron jobs / task schedulers – Use system schedulers (e.g., cron on Linux, Task Scheduler on Windows) to trigger API requests at set intervals.
ETL pipelines – Integrate API calls into your data ingestion workflows (e.g., Airflow, Prefect, or other orchestraon tools).
Cloud functions – Run serverless functions (AWS Lambda, Azure Functions, GCP Cloud Functions) to pull data and store it in your database or data warehouse.
Make sure your requests respect authentication requirements. For high-volume or time-sensitive pulls, batching requests and implementing retry logic is recommended.
3.6 What do the confidence levels mean?
Like other prediction products, our model computes a confidence score ranging from 0 (no confidence) to 4 (high confidence). For shadow price predictions, confidence is calculated based on historical tendencies and the prediction's proximity to zero.
3.7 Can I reproduce or resell this data?
No, you cannot reproduce this data for third-party use or resale purposes.