Home / Libraries / ML4T Data / Docs
ML4T Data
ML4T Data Documentation
Unified market data acquisition from 19+ providers
Source on GitHub
Skip to content

Databento Provider

Provider: DataBentoProvider Website: databento.com API Key: Required Free Tier: $125 credit

Overview

Databento provides institutional-grade market data across 45+ exchanges with 15+ years of history. DataBentoProvider exposes OHLCV-oriented historical fetches, continuous futures helpers, OPRA option helpers, schema discovery, and direct access to the native databento.Historical client for advanced workflows.

Best For: Professional futures research, institutional-quality data

Pricing: | Tier | Price | Features | |------|-------|----------| | Metadata and symbology | Free | Discovery, cost estimates, symbol resolution | | Usage-based | Pay as you go | Historical time-series data billed by volume | | Subscription | See current pricing | Bundled historical and live access |

Quick Start

import os
os.environ["DATABENTO_API_KEY"] = "your_key_here"

from ml4t.data.providers import DataBentoProvider

provider = DataBentoProvider()

# These authenticated discovery calls are not metered.
datasets = provider.get_available_datasets()
schemas = provider.get_available_schemas("GLBX.MDP3")

# Historical time-series calls are metered.
df = provider.fetch_ohlcv("ES", "2024-01-01", "2024-06-01", frequency="daily")

# Multiple schemas for one symbol
df = provider.fetch_multiple_schemas(
    symbol="ES",
    start="2024-01-01",
    end="2024-06-01",
    schemas=["ohlcv-1d", "ohlcv-1h"],
)

OPRA Options

The wrapper provides a small OPRA workflow for listed-options research:

  1. Estimate request cost before downloading broad quote or bar data.
  2. Discover option contracts from Databento definitions.
  3. Fetch OHLCV bars or consolidated OPRA quotes for selected contracts.
from ml4t.data.providers import DataBentoProvider

provider = DataBentoProvider()

estimate = provider.estimate_opra_cost(
    symbols=["SPY   240119C00480000"],
    start="2024-01-02",
    end="2024-01-02",
    schema="cbbo-1m",
)

chain = provider.fetch_option_chain(
    underlying="SPY",
    session_date="2024-01-02",
    expiry="2024-01-19",
    right="call",
    min_strike=440,
    max_strike=500,
)

bars = provider.fetch_option_ohlcv(
    contract="SPY   240119C00480000",
    start="2024-01-02",
    end="2024-01-02",
    frequency="daily",
)

quotes = provider.fetch_option_quotes(
    contract="SPY   240119C00480000",
    start="2024-01-02",
    end="2024-01-02",
    schema="cbbo-1m",
)

fetch_option_quotes() keeps Databento's consolidated OPRA publisher when publisher_id is present. Set consolidated_only=False to retain publisher-level quote rows.

Supported Schemas

Schema Description Use Case
ohlcv-1d Daily OHLCV End-of-day analysis
ohlcv-1h Hourly OHLCV Intraday patterns
ohlcv-1m Minute OHLCV Short-term strategies
trades Tick trades Microstructure
cbbo-1m OPRA consolidated quotes Listed-options quotes
mbp-10 10-level depth Order book analysis via native SDK

Exchange Coverage

Category Exchanges
Equity Index CME (ES, NQ, YM, RTY)
Energy NYMEX (CL, NG, HO, RB)
Metals COMEX (GC, SI, HG)
Rates CBOT (ZN, ZB, ZF, ZT)
FX CME (6E, 6J, 6B, 6A)
Agriculture CBOT (ZC, ZW, ZS)

45+ exchanges, 650,000+ symbols, 15+ years history.

Continuous Futures

# Fetch continuous front-month contract
df = provider.fetch_continuous_futures(
    root_symbol="ES",
    start="2020-01-01",
    end="2024-12-01",
    frequency="daily",
    version=0,
)

API Key Setup

# Environment variable
export DATABENTO_API_KEY=your_api_key_here

Get your API key at databento.com.

Cost Control

Databento documents metadata, symbology, account management, and cost estimation as free. Historical time-series responses are metered by bytes delivered. Estimate a request before downloading data, and use account budgets to enforce a maximum spend. The ML4T release workflow calls only authenticated metadata methods and cannot incur time-series usage charges.

Advanced Workflows

Feature Priority Notes
MBO (Market by Order) LOW Full order book
WebSocket streaming NOT PLANNED Use native SDK
Symbology API LOW Symbol resolution

For advanced Databento schemas, batch jobs, symbology resolution, and live streaming, use provider.client directly.

See Also