Forecast tomorrow's price with AI-driven insights powered by deep learning.
All tickers in the database with their date ranges. Use the filters to narrow the view.
Enter a Yahoo Finance ticker symbol (e.g. GC=F, ^GSPC, AAPL) to train the model and store predictions. This may take a few minutes.
Remove all stored predictions for a ticker. This cannot be undone.
This app predicts next day prices using an AI model built with TensorFlow. It is an AI-powered tool designed to forecast the next day’s price based on historical market data and technical indicators.
It uses a sliding window approach to capture patterns from the past 100 trading days, transforming this data into a feature-rich input for a neural network built with TensorFlow. The model incorporates indicators like RSI, MACD, Bollinger Bands, EMA, and volume-based metrics to enhance its understanding of market momentum, volatility, and trend behaviour.
Once trained, the model predicts future prices and evaluates its performance using metrics such as MAE, RMSE, MAPE, and directional accuracy. These metrics help assess not just how close the predictions are, but whether the model correctly anticipates upward or downward movements.
Simulate how an investment would have performed if you followed the AI predictions strictly. After market close each day, the model predicts tomorrow's close. If it expects a rise, a market-on-open buy order is placed for the next morning. If it expects a drop, a sell order is placed. Trades execute at the next day's open price. On buy days the return is open-to-close; on days you already hold, the full close-to-close return (including the overnight gap) is captured; on sell days you capture the overnight return up to the open. The portfolio is either fully invested or fully in cash.
Access prediction data programmatically using the REST endpoints below. All responses are JSON.
Returns an array of all loaded ticker symbols with their date ranges and row counts.
Example response:
[
{
"symbol": "GC=F",
"min_date": "2020-01-02",
"max_date": "2026-03-26",
"total_rows": 1560,
"rows_with_real": 1559
}
]
Returns prediction data, accuracy metrics, and the next-day forecast for a given ticker symbol.
Query parameters:
| Parameter | Type | Required | Description |
|---|---|---|---|
symbol | string | Yes | Ticker symbol, e.g. GC=F |
days | integer | No | Limit results to the last N days |
start | string | No | Start date filter (YYYY-MM-DD) |
end | string | No | End date filter (YYYY-MM-DD) |
Example request:
Example response:
{
"correct_direction_pct": 57.14,
"mae": 12.35,
"rows": [
{
"Date": "2026-03-26",
"Symbol": "GC=F",
"Real_Close": 3050.20,
"Predicted_Close": 3045.80,
"Correct_Direction": true,
"Close_Correct": true
}
],
"next_day": {
"date": "2026-03-27",
"predicted_close": 3052.10
}
}
cURL:
curl "https://your-domain.com/api/data-summary" curl "https://nextdayprice.com/api/predictions?symbol=GC%3DF&days=30"
Python:
import requests
# List all tickers
tickers = requests.get("https://nextdayprice.com/api/data-summary").json()
print(tickers)
# Get predictions for a ticker
data = requests.get("https://nextdayprice.com/api/predictions",
params={"symbol": "GC=F", "days": 30}).json()
print(f"Direction accuracy: {data['correct_direction_pct']}%")
print(f"Next day forecast: ${data['next_day']['predicted_close']}")
JavaScript:
// List all tickers
const tickers = await fetch("/api/data-summary").then(r => r.json());
console.log(tickers);
// Get predictions
const data = await fetch("/api/predictions?symbol=GC%3DF&days=30")
.then(r => r.json());
console.log(`Next day: $${data.next_day.predicted_close}`);
For inquiries, reach out to info@nextdayprice.com
We respect your privacy. No personal data is stored or shared with third parties.