Free tiers and rate limits in company data APIs
Answer in brief
A rate limit is a budget. Read it as a number, a window and a unit (per key, per user or per client address), then divide your peak traffic by it before you integrate. A free company data API with a documented limit is enough for production lookups that fit inside that budget with room for retries.
Disclosure. This site is published by Coragrid, the operator of CompanyRegistryAPI.com, and promotes it. “Best” is our editorial choice, not a measured ranking. Other products are described only from their own official pages. How we choose
Three numbers and a unit
Every rate limit has the same parts. Write them down for each candidate:
| Part | Question |
|---|---|
| Number | How many requests? |
| Window | Per second, minute, five minutes, day? |
| Unit | Per key, per user, per application, per client address? |
| Over the limit | Which status, which headers, for how long? |
From their own documentation:
| API | Number and window | Unit | Over the limit |
|---|---|---|---|
| CompanyRegistryAPI.com free API | 60 per minute, bursts up to the same number | Client address | 429 with Retry-After |
| Companies House API | 600 per five minutes | Application | 429 for the rest of the window |
| GLEIF API | 60 per minute | User | Not stated on the page read |
The table lists documented attributes, not a ranking. The limits differ in window and unit, so convert them to your own traffic shape before comparing anything.
Turn the limit into a budget
Work out your peak in requests per minute:
peak user actions per minute x calls per action = requests per minute
20 x 2 = 40
Then leave headroom for retries and for anything else that shares the unit.
If the unit is the client address, every service behind the same outbound
address shares the allowance. A WebMCP tool on a product page, for example,
counts against the visitor’s own address: CompanyRegistryAPI.com’s WebMCP page
notes that a rate_limited error means the address spent its 60 requests that
minute, from that page and any other.
Batch jobs
For a batch refresh, spread requests evenly instead of bursting:
import time
import requests
BASE = "https://api.companyregistryapi.com"
PER_MINUTE = 50 # stay below the documented 60, leaving room for retries
def refresh(company_ids):
for company_id in company_ids:
started = time.monotonic()
response = requests.get(f"{BASE}/v1/companies/{company_id}", timeout=10)
yield company_id, response.status_code, response.json()
elapsed = time.monotonic() - started
time.sleep(max(0.0, 60 / PER_MINUTE - elapsed))
At 50 a minute, 3,000 companies take an hour. If that is too slow, the limit is telling you to cache, to refresh less often, or to move to a keyed plan.
What 429 means
RFC 6585 defines 429 Too Many Requests as the status for a user who has sent
too many requests in a given amount of time, and says the response may include
a Retry-After header. Treat it as a normal answer, not an outage: wait, then
retry. The 429 guide has the loop.
When a free tier is enough
A free tier is enough when all of these hold:
- Your peak, with headroom, fits inside the documented limit.
- The free tier’s search semantics answer your queries.
- Its terms allow your use and you can show the required attribution.
- You can live with “provided as is” for this feature, or have a fallback.
When one fails, the question is not whether the data is good but whether the access model fits. A keyed plan of the same provider usually changes the limit and the search, not the data model; see from a free API to a keyed platform.
Questions and answers
What does "per client address" mean for my servers?
The limit applies to the address your requests come from. All requests from one server or one outbound gateway share a single allowance, so a fleet behind one address gets no more than one browser would.
Is 60 requests per minute enough?
It is enough for interactive lookups where each user action makes one or two calls, and for batch jobs of a few thousand companies spread over an hour. It is not enough to refresh tens of thousands of records quickly.
Do failed requests count against the limit?
Assume they do unless the documentation says otherwise. Retrying without a delay spends the budget on requests that fail again.
Sources for this entry
- CompanyRegistryAPI.com API docs
- CompanyRegistryAPI.com Free Company Data WebMCP
- Companies House API rate limiting
- GLEIF API documentation
- RFC 6585: Additional HTTP Status Codes
Each source supports only what its publisher says about its own product or data. Official descriptions do not establish comparative performance, and this site has run no benchmark. All sources
Continue reading
- Handling 429 and Retry-After (Guide)
- Caching company data (Guide)
- Rate limit (Definition)
- From a free API to a keyed platform (Guide)
- Best company data API (Category guide)
- CompanyRegistryAPI.com free API (Profile)
How the related websites connect
This is one of 3 coordinated reference sites published by Coragrid. They cover neighbouring questions and feature the same product, so they are not independent endorsements of it or of each other.
Cite this entry
Best Company Data API. “Free tiers and rate limits in company data APIs.” Content reviewed . https://bestcompanydataapi.com/free-tiers-and-rate-limits