specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Databento
providerId: databento
created: '2026-07-11'
modified: '2026-07-11'
reconciled: false
tags:
- Market Data
- Financial Data
- Historical Market Data
- Trading
- Rate Limiting
- Quotas
description: >-
  Databento advertises unlimited API calls and downloads with no per-request
  fees, so throughput is governed by usage-based billing rather than a hard
  per-minute request cap. In practice, historical requests are bounded by
  concurrency limits and by how much data your account and credit/plan allow you
  to stream, and very large pulls are steered toward the asynchronous Batch API,
  which materializes flat files rather than streaming synchronously. Live data
  is a continuous TCP subscription rather than a request-rate surface; its
  volume is governed by the schemas and symbols you subscribe to. Specific
  numeric per-account request or concurrency limits are not prominently
  published.
notes: >-
  No fixed numeric per-endpoint request-rate limit is published as of the review
  date. Use metadata.get_record_count, metadata.get_billable_size, and
  metadata.get_cost to size a query before running it, and prefer batch.submit_job
  for large historical extracts. Verify concurrency and any account-level limits
  with Databento support during reconciliation.
sources:
- https://databento.com/pricing
- https://databento.com/docs/api-reference-historical
- https://databento.com/docs/api-reference-historical/batch/batch-submit-job
- https://databento.com/docs/api-reference-live
responseCodes:
  throttled: 429
limits:
- name: Historical API Calls
  scope: account
  metric: requests
  limit: unlimited (no per-request fee)
  notes: Databento advertises unlimited API calls; no fixed numeric request-rate cap is published.
- name: Streaming Data Volume
  scope: account
  metric: bytes
  limit: usage-based (per credit / plan)
  notes: Bounded by billing - per-byte pay-as-you-go, free credit, or subscription depth - not by a request cap.
- name: Synchronous Query Size
  scope: request
  metric: records
  limit: steer large pulls to Batch
  notes: Very large historical extracts should use the asynchronous Batch API rather than synchronous timeseries.get_range.
- name: Batch Download Retention
  scope: job
  metric: days
  limit: 30
  notes: Batch output files remain re-downloadable for 30 days at no extra cost.
- name: Live Subscription
  scope: session
  metric: stream
  limit: continuous
  notes: Live is a persistent TCP/DBN subscription; volume is governed by subscribed symbols and schemas, not a request rate.
policies:
- name: Cost Estimation First
  description: Use metadata.get_record_count, metadata.get_billable_size, and metadata.get_cost to pre-size and pre-price a query before streaming.
- name: Prefer Batch for Bulk
  description: Route large historical extracts through batch.submit_job to produce downloadable flat files instead of long synchronous streams.
- name: Backoff Strategy
  description: Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com