Best 13F API in 2027: 7 Holdings APIs Compared
The best 13F API for institutional holdings data in 2027: 7 providers compared on 13F filings coverage, CUSIP scope, amendments, free tier and price.

Search for the best 13F API and page one is three vendors ranking themselves first. We sell one of these APIs too, StockFit, and we rank it first below, but we hold every claim to one rule: it comes from the vendor's own pricing page or docs as checked on 2026-11-18, or from a live response we reconciled against the SEC filing. Where a vendor does not publish something, the table says so instead of guessing.
This post also shows what none of the ranking pages do: a real 13F holdings row next to the information table it came from, reconciled to the share. If you only want the table, jump to the comparison. If you want to check any 13F API yourself, the reconciliation script works against every vendor here.
What 13F data is, and what it is not
Form 13F-HR is the quarterly report an institutional investment manager files once it holds at least $100 million in 13(f) securities on the last trading day of any month in a calendar year. The filing has a cover page (manager name, CIK, report period, other managers included) and an information table: one line per position, with the issuer CUSIP, title of class, reported value, shares or principal amount, put or call, investment discretion and voting authority (sole, shared, none). The 13F filing deadline is 45 days after quarter end, so the Q3 2026 13F filings were due November 16, 2026.
Three limits matter before you pick any 13F data API. 13F holdings are long positions only: the SEC's Form 13F FAQ says not to report short positions and not to net them against longs. Options appear only as long puts and calls, on their own lines. And managers can ask for confidential treatment of some positions, which then show up in a later filing or never. 13F data tells you what a manager owned on the last day of the quarter, disclosed up to 45 days later, not what it owns today.
That 45-day lag is also why 13F data is never complete on the deadline. In our own data, about one in five managers files on the last allowed day. We covered how far a quarter fills in each day of the window in our 13F filing data lag study.
How to choose a 13F API: six tests
Every 13F API reads the same SEC filings, so the differences are in what each one does with them. Six tests separate them, and you can run each in minutes against a free key or a trial.
- Issuer CUSIP scope. Ask for the institutional holders of GOOGL. Do you get Class A holders only, or every Alphabet class summed together? An API that keys on the issuer instead of the ticker CUSIP double counts dual-class companies and ETF trusts.
- Amendment handling. A 13F-HR/A is either a restatement (it replaces the original) or a new-holdings amendment (it adds lines to it). An API that treats every amendment as a replacement loses positions. One that adds every amendment double counts them.
- Quarter completeness. On day 30 of the filing window a quarter is half filed. Does the API tell you, or does it serve the half-filed quarter as the latest?
- Holdings deltas. Does each row carry the prior quarter's shares, or do you join two quarters yourself? The details of a good change feed are in our post on detecting institutional position changes.
- History back to which quarter. 13F information tables became structured XML in 2013. Anything older is text, and few vendors parse it.
- An accession number on every row. Without it, you cannot open the public filing behind a number and check it.
And a seventh test that decides most real projects: does the same key cover the rest of the ownership picture? 13F is one of three ownership filings. Schedules 13D and 13G name the 5% owners, and Forms 3, 4 and 5 report the insiders. Most products that use 13F data end up needing all three, plus the company's financials.
The best 13F APIs in 2027 at a glance
Seven providers, compared on the tests above. Prices and limits come from each vendor's own pricing page or docs, checked on 2026-11-18. A yellow mark means the vendor does not document the behavior, not that it is missing.
| Provider | 13F history | Free access | Entry plan with 13F | Accession per row | Amendments merged | Same key also covers |
|---|---|---|---|---|---|---|
| StockFit | Holdings from Q2 2013 | 50 premium calls on any free key | Professional (see pricing) | 13D/13G, Form 4, fundamentals, ETF holdings | ||
| sec-api.io | Holdings from Q1 2013, filing metadata from 1998 | 100 calls | $49/mo personal, $199/mo business | Form 3/4/5, N-PORT; 13D/13G from business plan | ||
| Arkolith | Not published | $1 trial, 3 days | $29/mo Starter | Form 4 | ||
| Financial Datasets | 5+ years (endpoint docs) | None ($20 buys 1,000 calls) | $200/mo Personal | 13D/13G, insider trades, fundamentals | ||
| Financial Modeling Prep | Not published | No 13F on the free plan | $139/mo Ultimate ($99/mo annual) | Fundamentals, prices, transcripts | ||
| WhaleWisdom | Quarters from Q1 2001 | Last 8 quarters, current excluded | Subscription (see their pricing) | Backtester, research UI | ||
| SEC EDGAR | Every 13F ever filed | Free, 10 requests/second | Free | Every filing, unparsed |
- 13F history
- Holdings from Q2 2013
- Free access
- 50 premium calls on any free key
- Entry plan
- Professional (see pricing)
- Accession
- Yes
- Amendments
- Yes
- Also covers
- 13D/13G, Form 4, fundamentals, ETF holdings
- 13F history
- Holdings from Q1 2013, filing metadata from 1998
- Free access
- 100 calls
- Entry plan
- $49/mo personal, $199/mo business
- Accession
- Partial or not documented
- Amendments
- Partial or not documented
- Also covers
- Form 3/4/5, N-PORT; 13D/13G from business plan
- 13F history
- Not published
- Free access
- $1 trial, 3 days
- Entry plan
- $29/mo Starter
- Accession
- Yes
- Amendments
- Yes
- Also covers
- Form 4
- 13F history
- 5+ years (endpoint docs)
- Free access
- None ($20 buys 1,000 calls)
- Entry plan
- $200/mo Personal
- Accession
- Yes
- Amendments
- Partial or not documented
- Also covers
- 13D/13G, insider trades, fundamentals
- 13F history
- Not published
- Free access
- No 13F on the free plan
- Entry plan
- $139/mo Ultimate ($99/mo annual)
- Accession
- Partial or not documented
- Amendments
- Partial or not documented
- Also covers
- Fundamentals, prices, transcripts
- 13F history
- Quarters from Q1 2001
- Free access
- Last 8 quarters, current excluded
- Entry plan
- Subscription (see their pricing)
- Accession
- Partial or not documented
- Amendments
- Partial or not documented
- Also covers
- Backtester, research UI
- 13F history
- Every 13F ever filed
- Free access
- Free, 10 requests/second
- Entry plan
- Free
- Accession
- Yes
- Amendments
- No, you build it
- Also covers
- Every filing, unparsed
Notes: StockFit's ownership endpoints are on the Professional plan, and every free account includes 50 lifetime calls to endpoints outside its plan, enough to run the tests above. sec-api.io returns the accession number on the filing object that wraps each holdings array, not on each holding, and an amendmentType of RESTATEMENT or NEW HOLDINGS on each filing, with merging left to you. Financial Datasets' pricing page mentions 30+ years of core history, while its institutional holdings endpoint docs state 5+ years. FMP lists Form 13F as “No Coverage” on its Basic, Starter and Premium plans. WhaleWisdom's API docs limit calls to 20 per minute. Also worth a look and not ranked here: HoldingsIntel (a free API tier of 200 calls a day, 13F history from 2014 on its Pro+ plan) and edgar.tools (13F endpoints on its Professional plan).
1. StockFit, best for the whole ownership stack under one key
StockFit's 13F API is nine ownership endpoints next to the insider, filings and fundamentals endpoints, all on one key. /api/ownership/institutional-holders answers who owns a stock, /api/ownership/portfolio answers what a manager owns, and /api/ownership/institutional-holders/history gives the quarterly holder count, shares and value. /api/ownership/summary rolls a stock's institutional ownership, top holders, beneficial owners and insiders into one call. Here is Apple's largest holder for the June 2026 quarter, an unedited row from production:
curl -H "Authorization: Bearer $KEY" \
"https://api.stockfit.io/v1/api/ownership/institutional-holders?\
symbol=AAPL&reportDate=2026-06-30&pageSize=1"{
"page": 1,
"pageSize": 1,
"totalResults": 6993,
"data": [
{
"managerName": "BlackRock, Inc.",
"managerCik": 2012383,
"value": 336524794350,
"shares": 1162996939,
"shareType": "SH",
"investmentDiscretion": "SOLE",
"soleVoting": 1048186191,
"sharedVoting": 0,
"noneVoting": 114810748,
"reportDate": "2026-06-30",
"prevShares": 1144677448,
"prevValue": 290507689476,
"pctOfShares": 7.96,
"accessionNumber": "0002012383-26-003238",
"prevPctOfShares": 7.84,
"putCall": null,
"cusip": "037833100",
"symbol": "AAPL"
}
]
}Each row carries the prior quarter (prevShares, prevValue), the voting authority split, the issuer CUSIP and the accession number of the 13F filing it came from. How each test from above comes out:
- CUSIP scope. A
symbol,cusipor FIGI lookup returns holders of that one security. BlackRock's June 2026 filing shows up as 460,972,803 GOOGL shares under02079K305and, in a separate GOOG call, 369,922,502 shares under02079K107. Aciklookup returns every class of the issuer, one row per manager and CUSIP. - Amendments. Restatements replace the original filing and new-holdings amendments are added to it, whatever order they reach EDGAR in.
- Completeness. A quarter becomes the default 50 days after it ends, once the 45-day 13F filing deadline and weekend shifts have passed. Pass
reportDateto pin any quarter, including one still being filed. - Options. Put and call lines stay in their own rows with
putCallset, so a call position never inflates a share count. - History. Full holdings from the June 2013 quarter, when information tables became XML.
The holder history is useful on its own. Apple's count steps up in every December quarter and stays flat in between. That is the 13F rule at work, not a data glitch: a manager that crosses $100 million during a year files its first 13F for the quarter ending December 31.
Then comes the part a 13F-only vendor cannot do. With the same key, /api/ownership/beneficial-owners returns the 5% owners from Schedules 13D and 13G (see our 13D and 13G API guide), /api/insider-transactions returns deduplicated Form 4 trades, and the financial statements come from the same SEC filings with their accession numbers. Institutional holders, 5% owners, insiders and financials share one auth, one schema style and one bill. More on that bundle on the 13F data API for hedge funds page.
2. sec-api.io, best for raw 13F filings at volume
sec-api.io is a filings API first, and its Form 13F API mirrors the filing closely. It covers 13F holdings from Q1 2013 and 13F filing details back to 1998 without holdings, includes managers that stopped filing, and says it indexes a new 13F about 300 milliseconds after it is published. You query with a POST search syntax over holdings or cover pages, and bulk JSONL downloads cover whole datasets. Each holding has the issuer CUSIP, ticker, value, shares or principal amount, put or call, investment discretion and voting authority, and the accession number sits on the filing object around the holdings array.
The tradeoff is that you get filings, not positions. Each filing reports an amendmentType, and turning an original plus its amendments into one view of the quarter is your job. Plans are $49 a month for individuals and small startups and $199 a month for internal business use, with 100 free calls to start. 13D and 13G are not on the personal plan.
3. Arkolith, best for agent workflows on 13F and Form 4
Arkolith is a narrow, deep 13F holdings API built for AI agents: resolved holdings, quarter deltas, point-in-time reads and an MCP server, plus Form 4 insider activity. Its docs say every holding row is tied to an SEC accession and that superseded filings are modeled so corrections do not double count. That is the same audit trail argument we make, and it holds for them too.
Pricing is credit based: a $1 three-day trial with 300 credits, then $29, $79 or $149 a month. Arkolith says itself that it is not a fundamentals or price feed, and it does not publish how many quarters of history it holds. If your agent only needs 13F and Form 4, it is a strong fit. If it also needs financials or 13D/13G, you add a second vendor.
4. Financial Datasets, closest in breadth
Financial Datasets is the competitor closest to a whole-stack answer: financial statements, insider trades, institutional holdings, 13D/13G holders and SEC filings under one API. Its institutional holdings endpoint takes a ticker, or a filer CIK on the by-investor endpoint, returns accession_number and put_call per row, and by default shows only filers whose latest 13F still holds the stock. Historical holders need report_period filters.
The endpoint docs state 5+ years of history for 13F holdings. Pricing is per request: $20 one time for 1,000 requests, $200 a month for 100,000 on a personal license, and $1,000 a month for commercial use. Amendment handling is not documented.
5. Financial Modeling Prep, only if you already need Ultimate
FMP has a broad Form 13F surface: filings extracts, holder performance, industry breakdowns and a positions summary. Its pricing page lists Form 13F as “No Coverage” on the Basic, Starter and Premium plans, so 13F data starts at Ultimate, $139 a month or $99 a month billed annually. If you already pay for Ultimate for global coverage or transcripts, the 13F endpoints come with it. FMP does not document amendment handling or an accession number on its 13F rows. FMP's fundamentals are normalized and drop the filing-level audit trail, as covered in our fundamentals API comparison.
6. WhaleWisdom API, best for research UI plus backtester
WhaleWisdom is a 13F research site first, with an API on top. Its quarter list starts with Q1 2001, and commands cover stock and filer lookups, holders, holdings comparisons and its backtester. Without a subscription the API serves only the last eight quarters, not including the current one. Every account is limited to 20 requests a minute, so it suits research pulls and exports better than a production backend. Teams looking for a WhaleWisdom API alternative usually want higher limits, current-quarter data, or 13F next to other datasets.
7. SEC EDGAR, the free 13F baseline
Every API above reads SEC EDGAR, and you can too, free, at 10 requests a second. A manager's submissions JSON lists its 13F-HR filings, and each filing holds a primary_doc.xml cover page and an information table XML. That is the free 13F holdings API. What you build yourself: CUSIP-to-ticker mapping, the amendment merge, prior-quarter deltas, the completeness rule, and a way to search holders across thousands of filers. It is a real option for one-off research, and the next section shows the reading part takes about 20 lines.
Checking a 13F API against the filing: BlackRock's Apple position
A 13F API is only as good as its reconciliation. BlackRock's June 2026 13F-HR (accession 0002012383-26-003238) has 49,968 information table entries. This script reads them straight from EDGAR and adds up the Apple lines:
import requests
import xml.etree.ElementTree as ET
UA = {"User-Agent": "your-app/1.0 (you@example.com)"}
url = ("https://www.sec.gov/Archives/edgar/data/2012383/"
"000201238326003238/form13fInfoTable.xml")
root = ET.fromstring(requests.get(url, headers=UA, timeout=120).content)
ns = {"t": root.tag.split("}")[0].strip("{")}
shares = calls = 0
for row in root.findall("t:infoTable", ns):
if row.findtext("t:cusip", namespaces=ns) != "037833100":
continue
amount = int(row.findtext("t:shrsOrPrnAmt/t:sshPrnamt", namespaces=ns))
if row.findtext("t:putCall", namespaces=ns):
calls += amount
else:
shares += amount
print(shares, calls) # 1162996939 1661200BlackRock reports Apple on 34 lines. 24 share lines add up to 1,162,996,939 shares worth $336,524,794,350, exactly the shares and value in the API row above. The other 10 lines are call options on 1,661,200 shares, which the API keeps in a separate row with putCall set. The same check on Berkshire Hathaway's June 2026 filing gives 12 Apple lines and 227,917,808 shares worth $65,950,296,923, matching /api/ownership/summary to the share and the dollar.
Which 13F API should you choose?
- You need 13F next to 13D/13G, Form 4 and fundamentals, with an accession number on every row: StockFit.
- You want raw 13F filings with real-time indexing and bulk downloads, and will merge amendments yourself: sec-api.io.
- Your AI agent only needs 13F and Form 4: Arkolith.
- You want a broad dataset priced per request: Financial Datasets.
- You already pay for FMP Ultimate: use its 13F endpoints before adding a vendor.
- You want a 13F research site with a backtester and occasional API pulls: WhaleWisdom.
- You will build the parsing and the amendment merge yourself: SEC EDGAR.
To run the six tests against StockFit, : its 50 premium calls reach the ownership endpoints. If insider data is the bigger part of your project, our insider trading API comparison runs the same exercise for Form 4. This is data, not investment advice.
FAQ
Q.What is the best API for 13F data?
Q.Is there a free 13F API?
Q.How often is 13F data updated?
Q.Does 13F data include short positions?
Q.Why do GOOG and GOOGL show different institutional holders?
Q.How do 13F amendments work in an API?
Q.How far back does 13F holdings data go?
Ready to build?
Free API key, no credit card.