Documentation

Welcome to
Bukuni Forecast

The AI forecasting API for production teams. Send your historical time series data, receive structured predictions — in a single REST call.

Quick Start Interactive API Reference →

Quick Start

Get your first forecast in under 2 minutes.

Authentication

API key setup and request headers.

Models

Understand the seven available models.

Frequencies

Daily, weekly, monthly and more.

Multiple Series

Forecast many series in a single API call.

API Reference

Interactive Swagger explorer for all endpoints.

Quick Start

Your first forecast in 60 seconds

Sign up for a free trial key, then POST your historical data.

1 Get your API key

Create a free account to receive your bk_... API key.

Create free account →

2 POST to /forecast cURL

curl -X POST https://forecast.bukuni.com/forecast \
  -H "X-API-Key: bk_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "data": [
      { "ds": "2023-01-01", "y": 1380 },
      { "ds": "2023-02-01", "y": 1420 },
      { "ds": "2023-03-01", "y": 1480 }
    ],
    "horizon": 3,
    "model":   "bukuni-auto",
    "freq":    "MS"
  }'

3 Receive your forecast 200 OK

{
  "forecasts": [
    { "unique_id": "series_1", "ds": "2023-04-01", "forecast": 1531.2 },
    { "unique_id": "series_1", "ds": "2023-05-01", "forecast": 1574.8 },
    { "unique_id": "series_1", "ds": "2023-06-01", "forecast": 1612.1 }
  ],
  "model":        "bukuni-auto",
  "horizon":      3,
  "series_count": 1
}

Authentication

API Key

Every request must include your API key in the X-API-Key header.

Required header — every request

X-API-Key: bk_your_api_key_here

Free Trial

100 calls · 14 days

Pro

1,000 calls · ₹9,999/mo

Pro+

10,000 calls · ₹79,999/mo

Create account → · View all plans →

Models

Choosing a model

Pass the model field in your request. Leave it empty to use bukuni-auto by default.

Model	Min data	Speed	Best for
bukuni-auto default	2 points	40–90s	Maximum accuracy, auto-adapts
bukuni-fast	2 points	5–10s	Short series, fast turnaround
bukuni-classic	2 points	<2s	Linear & seasonal patterns
bukuni-trend	2x horizon	30–60s	Trend + seasonality decomposition
bukuni-long	2x horizon	30–60s	Long forecast horizons
bukuni-multi	2x horizon	30–60s	Multiple correlated series
bukuni-deep	2x horizon	60–90s	Complex multi-pattern series

Frequencies

Setting the `freq` parameter

The freq value must match the actual gap between your data points.

MS

Monthly (default)

2024-01-01, 2024-02-01, …

D

Daily

2024-01-01, 2024-01-02, …

W

Weekly

Every 7 days

QS

Quarterly

2024-01-01, 2024-04-01, …

H

Hourly

2024-01-01 00:00, 01:00, …

Y

Yearly

2020-01-01, 2021-01-01, …

Multiple Series

Batch forecasting

Forecast any number of series in a single API call by using different unique_id values. The unique_id field is optional for single-series requests.

Two products, one request POST /forecast

{
  "data": [
    { "unique_id": "product_A", "ds": "2024-01-01", "y": 500 },
    { "unique_id": "product_A", "ds": "2024-02-01", "y": 530 },
    { "unique_id": "product_B", "ds": "2024-01-01", "y": 800 },
    { "unique_id": "product_B", "ds": "2024-02-01", "y": 850 }
  ],
  "horizon": 3,
  "model":   "bukuni-fast",
  "freq":    "MS"
}