addyAIMLprojects/btcusdt_spot_1m_05_2021_to_03_2026
BTCUSDT Spot 1-Minute OHLCV (May 2021 - Mar 2026) Overview 1-minute OHLCV candlestick data for the BTC/USDT spot pair on Binance, covering May 1, 2021 to February 28, 2026. Rows: 2,541,600 Completeness: 100.00% Sources Period Source Notes Full dataset Binance Data Collection Monthly kline ZIPs 2021-08-13 02:00-06:29 Bybit API 270 bars filled from Bybit BTCUSDT spot (Binance maintenance) 2021-09-29 07:00-08:59 Bybit API 120 bars… See the full description on the dataset page: https://huggingface.co/datasets/addyAIMLprojects/btcusdt_spot_1m_05_2021_to_03_2026.
BTCUSDT Spot 1-Minute OHLCV (May 2021 - Mar 2026)
Overview
1-minute OHLCV candlestick data for the BTC/USDT spot pair on Binance, covering May 1, 2021 to February 28, 2026.
- Rows: 2,541,600
- Completeness: 100.00%
Sources
Columns
Statistics
Data Quality
No gaps -- the dataset is 100% complete with no missing minutes.
Why start at May 2021?
Binance BTCUSDT 1-minute data is available from August 2017, but the earlier period (2017-2021) contains 31 maintenance gaps totaling 8,091 missing bars. Bybit minute-level data for BTCUSDT only starts from August 2021, making those earlier gaps unfillable. May 2021 is the earliest month after which all gaps can be backfilled from Bybit, providing a fully contiguous dataset.
Backfilled bars
470 bars across three Binance maintenance windows were sourced from Bybit BTCUSDT spot. Bybit prices closely track Binance but may differ by a few dollars.
Related datasets
- `Torch-Trade/btcusdt_spot_1m_03_2023_to_03_2026` -- shorter range (Mar 2023 - Feb 2026) but includes the Binance zero-fee promotion volume anomaly documentation.
Usage
from datasets import load_dataset
import pandas as pd
ds = load_dataset("Torch-Trade/btcusdt_spot_1m_05_2021_to_03_2026")
df = ds["train"].to_pandas()
df["timestamp"] = pd.to_datetime(df["timestamp"])
print(df.shape) # (2541600, 6)
print(df.head())License
MIT -- price data sourced from Binance and Bybit public data repositories.