Palantir Technologies Inc.

The full register, from 13F filings

The Register tab above is hand research: a dozen holders, each traced to the filing it came from. This asks the same question of the whole market for the quarter ended 31-mar-2026, where 2,945 managers reported a position. Broad where the research is deep, and shallow in the ways a 13F always is.

Built from 13F filings, which report US listed, long, US custodied positions only. A register from this source is a floor rather than a level: a holder's true position can be far larger, as BlackRock's ASML stake shows at 1.28m shares in a 13F against 26.3m in its 13G/A. Holdings held through derivatives, outside US custody, or below a manager's reporting obligation do not appear. Where a holder is marked as a family, its total sums several separate filing entities and describes something no single filing describes; the constituent entities and their share counts are listed so the total can be checked. That roll-up can legitimately exceed the family's own Schedule 13G, because a 13G is filed by one legal entity: Vanguard reports Alphabet through ten advisory entities, and The Vanguard Group Inc itself reported zero after its January 2026 disaggregation.

What a reader needs to know to read these numbers

• Palantir has three classes of common stock. As of 27 Jul 2026 (Q2 FY2026 10-Q cover page, filed 4 Aug 2026): 2,300,713,329 Class A shares (1 vote each), 101,340,151 Class B shares (10 votes each), and 1,005,000 Class F shares outstanding, for a combined total of 2,403,058,480 shares. Only Class A trades publicly as PLTR on Nasdaq. shares_outstanding in this file is that combined total.
• Class F mechanics, in plain language, from the Q2 FY2026 10-Q and the 24 Apr 2026 proxy: all 1,005,000 Class F shares are held in a voting trust established by the three Founders (Stephen Cohen, Alexander Karp and Peter Thiel), administered under a Founder Voting Trust Agreement with Wilmington Trust. Class F is a control mechanism, not a large economic stake: it is designed so that the Founders' Class F shares, together with Class A and Class B shares the Founders hold directly or have brought under a separate Founder Voting Agreement (and certain 'Designated Founders' Excluded Shares'), let the Founders control up to 49.999999% of the Company's total voting power, so long as the Founders collectively hold at least a minimum ownership threshold (100 million equity securities as of 30 Jun 2026). Each Class F share is convertible at any time, at the holder's option, into one Class B share.
• Combined founder voting power, its own basis, not converted to a share count for the register: the proxy's 'Founder Voting Control' table states that shares subject to the Founder Voting Agreement, the Founder Voting Trust and Designated Founders' Excluded Shares together give the Founders a 'Founder Total' of 49.999999% of total voting power (record date 6 Apr 2026), versus a combined 9.505% of all shares outstanding on an economic basis (75,304,406 Class A + 152,084,943 Class B + 1,005,000 Class F, which is the same total, 228,394,349 shares, as summing the three individual Karp, Thiel and Cohen rows in this register). This is the core story of Palantir's structure: the Founders hold under 10% of the equity but control just under half the vote.
• Individual founder rows in the register use the proxy's 'Shares Beneficially Owned' table (record date 6 Apr 2026) and are economic share counts only. The proxy's separate 'Percentage of Votes (Record Date Outstanding Shares)' column, which already weights Class B at 10 votes and (for the Founder Total row only) includes Class F, gives materially different numbers: Karp 11.4%, Thiel 7.9%, Cohen 3.0%, versus their economic pct_of_company of 3.90%, 4.09% and 1.51% respectively. Do not conflate the two.
• Institutional register rows are all Class A positions (CUSIP 69608A108), the only class that trades and the only class 13F filers can hold. pct_of_company in this file divides by the combined 2,403,058,480 shares across all three classes for consistency with the founder and officer rows; each institutional row's method_note also gives the percentage of Class A alone, which is 4 to 7 basis points higher in every case.
• BlackRock has two sourced figures that disagree because they are different dates and different filing types, both recorded rather than resolved: BlackRock's own Form 13F-HR for the quarter ended 30 Jun 2026 reports 189,817,693 Class A shares (used in the register), while Palantir's 24 Apr 2026 proxy names BlackRock as its only greater-than-5% stockholder at 158,217,849 Class A shares, sourced in turn to a Schedule 13G/A BlackRock filed 8 Nov 2024, which is 19 months older.
• Vanguard filed Form 13F-NT (notice, no holdings table) for the quarters ended 31 Mar 2026 and 30 Jun 2026 under CIK 0000102909, so its most recent 13F-HR with an actual PLTR position is for the quarter ended 31 Dec 2025, used in the register. A more current Vanguard position may exist but was not found within the time budget.
• Palantir itself filed Form 13F-HR as an institutional investment manager from 2022 through Q3 2023 (last filing 14 Nov 2023, period 30 Sep 2023), naming four small equity stakes it held in customer or counterparty companies: Allego N.V. (694,838 shares, $1.54m), MSP Recovery, Inc. (9,038,838 shares, $1.99m), Rubicon Technologies, Inc. (1,391,604 shares, $2.89m) and Surf Air Mobility Inc. (2,020,874 shares, $3.29m), totaling $9.72m. Palantir has not filed a 13F-HR or 13F-NT since. These positions are not carried into holdings here because they are roughly three years stale and at least one, Rubicon Technologies, was delisted from the NYSE in June 2024 and last traded on the OTC market in August 2024, so the 2023 figures would misrepresent Palantir's current holdings rather than describe them.
• The Q2 FY2026 10-Q (period 30 Jun 2026) reports two combined equity-investment balances, neither broken out by investee: publicly-traded equity securities of $184 million (up from $23.355 million at 31 Dec 2025), carried within marketable securities at fair value; and privately-held equity securities of $167 million (versus $170.0 million at 31 Dec 2025), carried in other assets under the measurement alternative (cost less impairment, adjusted for observable price changes, Level 3). The FY2025 10-K states the FY2025 privately-held figure rose from $64.9 million at 31 Dec 2024 to $170.0 million at 31 Dec 2025, with $30.7 million of upward observable-transaction adjustments during 2025, but names no investee in either filing.
• Palantir's 10-K and 10-Q risk factors describe an active but unnamed program: the Company discloses it 'may continue to make strategic investments' and has entered 'Investment Agreements' to purchase or commit to purchase shares of 'various entities, including special purpose acquisition companies and/or other privately-held or publicly-traded entities,' often in connection with customer contracts. No individual counterparty, share count, or dollar amount is disclosed for any current Investment Agreement in either filing searched.
• Market capitalization of $418.23 billion is the figure published by stockanalysis.com as of 14 Aug 2026, the same reference date used by other files in this database. Cross-check: 418.23bn divided by the 2,403,058,480 combined shares outstanding implies roughly $174 per share.
• Context only, not a register row: stockanalysis.com reports institutional ownership of Palantir at 60.93% and insider ownership at 7.41% of shares outstanding, as of 16 Aug 2026 (source: https://stockanalysis.com/stocks/pltr/statistics/). This is a third-party aggregate, not a filing, and it is a superset that already contains the seven named institutional holders in the register (Vanguard, BlackRock, State Street, Geode, FMR, Morgan Stanley and JPMorgan), so it is not carried as a register row to avoid any reader summing it against them. The page does not state whether the 60.93% figure includes Class B and Class F or is Class A only. The 7.41% insider figure is in the same range as the 9.87% this file computes from the proxy's officers-and-directors group, the difference likely reflecting measurement date and whether option or RSU-vested shares are counted.