Early access is open. Free sample datasets are live now. Download a sample

Rate Limits and Plan Quotas

Details

Docs navigation

Each plan has a request rate, and some datasets add their own daily or monthly quotas on the free tier. This page lists the planned API rate limits for market data on every plan, shows the response headers that tell you where you stand, and gives a retry pattern that will not get you locked out. You need it before you write a loop that pulls a few thousand symbols.

Early access. These are the planned limits for v1. They may change before launch, and we will post any change in the changelog.

Pricing

Limits by plan

Plan Price Requests per minute (default) Typical use
Free $0 5 Trying the API, small personal scripts
Starter $29/mo 60 Daily research, a few hundred symbols
Pro $79/mo 300 Intraday backtests, real-time dashboards
Business $249/mo 1,000 Small funds and apps, commercial internal use
Enterprise Quote Agreed per contract Redistribution, full depth

Limits apply per account, across all of its keys.

Datasets

Dataset-specific quotas

A few free-tier datasets use a daily or monthly cap instead of, or as well as, the per-minute rate:

Dataset Free tier quota Paid tiers
Option chain and options EOD 5 underlyings per day All underlyings from Starter
FX latest and daily rates 1,000 calls per day Starter: 60 calls/min for latest rates, 300 calls/min for daily and intraday history
Crypto prices and bars 10,000 calls per month Plan default per minute
Technical indicators Each call counts as one request Same

Batch calls help a lot here. The snapshot endpoint accepts up to 250 symbols in one request on every plan, and it counts as one call.

Pricing

History depth by plan

Rate is one limit; history is the other. A request outside your plan's window returns 403 plan_required rather than silently trimming the range.

Data Free Starter Pro Business
Daily stock bars 2 years Full Full Full
1-minute stock bars None 5 years, delayed Full, real time Full, SIP real time
Stock trades None None 10 years Full
Options EOD 1 year 5 years Full Full
FX 1-minute bars 1-hour bars, 30 days 5 years From 2010 From 2010
Crypto 1-minute bars 30 days 5 years Full Full

Keep reading

Reading the headers

Every response carries these headers:

Header Meaning
X-RateLimit-Limit Requests allowed in the current window
X-RateLimit-Remaining Requests left in the window
X-RateLimit-Reset Unix time (seconds) when the window resets
Retry-After Only on 429: seconds to wait before retrying

Details

Handling a 429

When you go over, you get HTTP 429 with error.code set to rate_limited. Wait the number of seconds in Retry-After, then try again. If you keep getting 429s, slow your loop down rather than retrying harder.

Example (Python):

import time
import requests

def get(url, headers, params=None, max_tries=5):
    for attempt in range(max_tries):
        r = requests.get(url, headers=headers, params=params)
        if r.status_code != 429:
            r.raise_for_status()
            return r.json()
        wait = int(r.headers.get("Retry-After", 2 ** attempt))
        time.sleep(wait)
    raise RuntimeError("still rate limited after retries")

A steady pace beats bursts. On Starter, one request a second (60 a minute) will never trip the limit.

Watch out

Notes and gotchas

  • WebSocket messages do not count against the REST limit. Opening a connection does count as one request.
  • Paginated calls count once per page. A 2-million-row pull at 50,000 rows a page is 40 requests.
  • A 429 response does not cost you anything extra, but repeated hammering after a 429 can lead to a temporary block on the key.
  • Free tier data is for personal use. Commercial internal use starts at Business.
  • If you need many symbols over a long range, a dataset download is usually cheaper and faster than thousands of API calls.

Keep reading

Next steps

Details

Was this page useful?

[Yes] [No]

Need a higher limit for a specific job? Tell us on the custom data form.