What Pi Network is, and what is still unverified about it

Pi Network is three things with three kinds of evidence behind them: a phone app that credits a balance to people who check in daily, a separate ledger run on the project’s own network, and a migration step, gated by an identity check, that carries balances from the first to the second. A question about what Pi is worth rests on all three. The project’s status has also changed several times and its own pages do not always keep up, so each status claim here carries the date it was read, 3 October 2026, and the page it comes from.

What the app does

The app asks for one action. Pi’s FAQ puts it as “To start mining Pi, check in every 24 hours and hit the lightning button”, and adds that once a session has started the app can be closed. The whitepaper defines the base role the same way: a Pioneer is a user “simply confirming that they are not a ‘robot’ on a daily basis”. Each check-in raises a balance that lives in the app, at a rate set by a formula. In Pi’s MiCA white paper, a 2025 filing under the EU’s markets regulation, the inputs are mining activity, Security Circles, app usage, node operation and lockups, and the base rate “is dynamically adjusted to align with the network’s supply limits and the total contributions from active Pioneers”. The paper does not say who makes the adjustment.

The whitepaper names four roles. Beyond the Pioneer there is the Contributor, who supplies “a list of pioneers he or she knows and trusts”, the Ambassador, who introduces other users, and the Node, a Pioneer who also runs the node software on a desktop or laptop. The FAQ adds a 25% bonus on the base rate “for each person you invite to the network” who is mining at the same time, and says the bonus “only applies to 1 level of referral relationship”. Pi’s March 2026 post on second migrations says referral bonuses migrate only for Referral Team members who have “fully passed KYC”, Pi’s identity check (know your customer).

None of this is mining in the proof-of-work sense. There, mining is the competitive hashing that produces blocks, and the coins paid to the winner are issued by a consensus rule (the tokenomics piece traces that rule in Bitcoin Core). Pi’s MiCA paper says its consensus design “does not rely on competing hashing power”, and the whitepaper’s section on mobile users says that to confirm a transaction the app “connects to one or more Nodes” to ask whether it was recorded. So “mining” names how a balance is earned in the app, through presence, a trust list, invitations, app use and optional lockups. It does not name how a ledger is secured, and a balance in the app is not an entry on that ledger. Other apps and cloud contracts sold as mining are examined in what cloud mining claims rest on.

Where the ledger is

Pi calls its ledger the Mainnet. The whitepaper says Pi’s consensus algorithm “builds atop” the Stellar Consensus Protocol, and the MiCA paper describes it as “built on a fork of the Stellar Consensus Protocol”. The ledger’s public API reports Stellar software. Its root document, read on 3 October 2026, names the network “Pi Network”, gives the versions as stellar-core 27.1.0 and Horizon 27.0.0, and listed ledger 29,031,256 as closed at 16:24:08 UTC that day. So there is a live ledger of the Stellar lineage, readable without an account.

That lineage turns the question from hashing to trust. In the Stellar protocol, per Stellar’s documentation, “each Core node decides on which other nodes it would like to trust to reach agreement”, so whether the result is independent depends on who the validators are and whom they trust.

On who runs the nodes, Pi’s Open Network launch post says “anyone can technically add nodes to the Mainnet blockchain, in terms of running the protocol and connecting to the network”, and that Pioneers will be invited gradually, with “initial preference for strong historical contributions and reliability scores”, to move nodes from Testnet to Mainnet. The node counts in the posts read are for Testnet (over 350,000, in the December 2025 review) or do not say which network (over 420,000 operators, in the Pi2Day 2026 recap). The pages read (the whitepaper, the MiCA paper, the FAQ, the node and roadmap pages, and the posts cited here) give no number of Mainnet validators, no names and no trust configuration. The MiCA paper says SocialChain, Inc. oversees “core protocol development”, which describes a role and not a validator set.

What the validators run is no better established than who they are. The one observation cited here is a build that Pi’s own API reports about itself: the commit it gives for stellar-core, 3589a696b0d4, is the commit upstream tags as v27.1.0 in the public stellar-core repository. A self-reported string shows what that endpoint says it runs, not how it was built or what the validators run. Pi’s node page says the blockchain component of the node “will be open sourced”, in the future tense, on a page that also describes the Open Network as upcoming, so the tense may be out of date. The MiCA paper describes the consensus layer as built on a fork, yet the commit reported is one that upstream itself tags, and neither document says how the two fit together. So the trust configuration is not the only thing the pages read leave unpublished: so is the source of what Pi’s validators run.

The ledger’s contents are public: accounts, balances and transactions can be read. The header of the latest ledger gives total_coins as 100,000,000,000, and so does the earliest ledger the API holds, ledger 2. That matches the whitepaper’s maximum supply. It records what exists in the ledger and not what has been distributed: the MiCA paper says the Foundation’s and liquidity pool’s tokens “were created from a genesis event”. Who holds the units, which are locked and how many circulate are not in the header. The API’s endpoints, listed in its root document, cover accounts, assets, claimable balances, ledgers, operations, payments and trades, and none totals or labels holdings.

KYC and the migration

Pi’s FAQ says its “native, decentralized KYC solution” validates the identity of Pioneers “before they are eligible to migrate their mined mobile Pi balance onto the Mainnet blockchain”. The whitepaper’s Mainnet phase says “only accounts validated to belong to distinct real individuals will be honored” and “balances of fake users or duplicate users will be discarded”. Until migration, then, the balance in the app is conditional on a verification the project runs. Pi’s April 2026 post on validator rewards says the work is done by human validators, 1,094,680 of them across 526,970,631 validations, alongside AI processing, and paid from a pool into which “each Pioneer who migrates to the Mainnet pays 1 Pi”.

Migration also had a deadline. In July 2024 Pi set a six-month grace period for applying and migrating, and said Pioneers who did not complete the steps “will forfeit most of their past Pi balance other than the Pi mined within the last 6 months before their Pi is migrated”. Deadlines were extended several times, and a 28 February 2025 post set 8:00 UTC on 14 March 2025 as the final extension, “now that Open Network has launched”. No Pi post cited here gives a figure for how much balance the rule removed.

“Enclosed” was the phase in which the Mainnet ran behind a firewall. Pi’s roadmap defines it as “live but with a firewall that prevents any unwanted external connectivity”. Pi’s anniversary post dates the Open Network launch to 20 February 2025, and the launch post says the firewall “has been removed”. What opened is connectivity, not participation without conditions: the launch post also says Mainnet participation requires KYC for Pioneers and KYB (know your business) for businesses. As of 3 October 2026 three Pi pages still describe the Open Network as to come. The roadmap says the Open Network period “will begin depending on the maturity of the Enclosed Network ecosystem”, the FAQ says the firewall “will be removed”, and the node page speaks of “the upcoming Open Network period”. The whitepaper carries its own caveat: its original March 2019 text “may need updates”. The launch is documented on the blog and not on those pages.

Pi’s counts show how far migration has gone. On 3 April 2026 Pi reported over 16.5 million people “KYC’d and migrated” and 18 million identities verified. On 17 September 2026 it described two groups held up: over 417,000 whose accounts had been flagged as possible duplicates and could now continue in KYC, and 497,000 who could not claim migrated balances because wallets created through its Fast-Track route held too little Pi for the gas fee. The post said the second group would get an update within a week. As of 3 October, 16 days after the post, none of the posts read says whether it shipped. Against those counts, a May 2026 post speaks of “over 60 million Engaged Pioneers”. The figures are weeks apart, both are lower bounds, and the May post does not define “Engaged”, so no share can be computed from them.

Supply and allocations

The whitepaper states a maximum total supply of 100 billion Pi: 80 billion eventually for the community (65 billion for Pioneer mining rewards, 10 billion for ecosystem building, to be managed in future by a non-profit foundation, and 5 billion for liquidity pools) and 20 billion for the Core Team. The Core Team’s allocation “gets unlocked at the same pace as the community progressively mines more and more Pi”, possibly with further lockup under “a self-imposed mandate”. Pi’s April 2025 tokenomics post turns that into a formula: every allocation tracks the community’s migrated mining rewards, so the “Effective Total Supply” is the migrated rewards divided by 65%.

The tokenomics piece asks what enforces each supply figure. Here the answer is a document and a formula: all 100 billion units are in the ledger from the start, and nothing in the API’s endpoints shows a release schedule, so the pace Pi describes is a statement and not something the ledger can be seen enforcing. The pages read do not list the addresses behind the Core Team, Foundation or liquidity allocations. The MiCA paper gives up to 10 billion for the Foundation and 5 billion for the liquidity pool under “Issuer Retained Crypto-Assets”, and describes core contributors’ tokens as “allocated tokens with vesting schedules”, with no amount or schedule. The paper also says two things about distribution. Its summary says the tokens “were initially distributed through a combination of initial allocations to contributors”, without saying whether it means the core contributors, while its Part A, on the person seeking admission, says “There was no ICO; tokens were only distributed through mining in the app”. Read together they leave open how many tokens contributors received outside mining.

The April 2025 post links Pi’s explorer as the source of migrated mining rewards. The explorer’s page script fetches its supply panel from socialchain.app, the backend named in its own configuration, at an address that carries the name of SocialChain, Inc., which the MiCA paper names as operator of Pi’s apps. That feed, stamped 16:23 UTC on 3 October 2026, gave about 11.24 billion Pi of “Migrated Mining Rewards”, of which about 6.11 billion is labelled “Currently Locked Mining Rewards” and, by the explorer’s own subtraction, about 5.14 billion unlocked. It gave a “Circulating Supply” of the same 11.24 billion, an “Effective Total Supply (R/65%)” of about 17.30 billion and, typed into the script and not read from any feed, a “Max Supply” of 100 billion. Two details govern how to read that. The circulating row equals the migrated rewards, so it counts the locked part (computed: about 54% of the migrated total). And the maximum is a constant.

Pi’s own proportions applied to the 11.24 billion give about 3.46 billion for the Core Team, 1.73 billion for the Foundation and 0.86 billion for liquidity (computed: 20, 10 and 5 sixty-fifths of the migrated rewards). Those are the amounts Pi’s April 2025 formula counts towards the Effective Total Supply, not balances the ledger shows as held or locked: with the migrated rewards they sum, before rounding, to the explorer’s 17.30 billion (computed). The MiCA paper, dated October 2025, said approximately 8.2 billion Pi were in circulation, without saying how circulation was counted, so the gap to the explorer’s 11.24 billion cannot be assigned to growth since then or to a different definition.

Listings, the filing and the money

Pi’s own list of approved businesses is its KYB list, an undated page of twelve names with no categories. Seven are spot exchanges (Kraken, OKX, Bitget, Gate, Pionex, LBank and MEXC), and the rest are Onramp.money, Onramper, Zypto, TransFi and Banxa. The page says only KYB-verified businesses can hold Pi Mainnet wallets, and the launch post tells Pioneers to avoid services that claim to integrate Pi but are not on it. Listing is a fact that can be checked at the venue: on 3 October 2026 the public market-data interface of each of the seven exchanges showed a PI spot market in its own records. That says a venue offers a market. How usable the market is at size depends on depth, not on presence.

The MiCA white paper is version 1.1, October 2025, prepared and filed by PiBit Ltd, a British Virgin Islands company registered on 24 December 2024 and wholly owned by the Pi Foundation, a Cayman Islands foundation. It gives 28 November 2025 as the start of admission to trading in the EU and EEA, initially on OKCoin Europe (trading as OKX), and describes PiBit as having “no (or minimal) full-time staff”. ESMA’s interim register of white papers for this category, in its OTHER.csv file as read on 3 October 2026, lists PiBit Ltd with the same document, the Malta Financial Services Authority as competent authority and a last-update date of 11 December 2025.

For a token of this kind a white paper is notified, not approved. MiCA Article 8(3) says competent authorities “shall not require prior approval”, and Article 6 requires every paper to carry fixed statements: that it “has not been approved by any competent authority” (6(3)), and that the asset may lose its value in part or in full, may not always be transferable or liquid, and is not covered by investor-compensation or deposit-guarantee schemes (6(5)). Pi’s paper has them because the regulation requires them, not because they are findings about Pi. What is specific to Pi is its statement that “there are no rights or obligations associated with the Pi token”, with no contractual, governance, dividend or redemption rights against the issuer or its affiliates, and its answer of “False” to the technology section’s audit question.

How the project earns is given in pieces. The Pi Ad Network, announced in March 2024, says “advertisers ultimately need to acquire and use Pi to place ads in the Pi ecosystem”, with Pi “distributed to the developers of Pi Apps displaying ads”. An App Studio option added in January 2026 lets creators deploy apps by watching ads, and Pi said that ad revenue “does not cover (not even on the same magnitude) the cost of app-generation” and is “subject to change or cancellation”. Pi also priced AI-powered app creation without a markup on the underlying AI costs (August 2026) and invited AI companies to contact it about the verified-human work done through KYC validation (April 2026). In the posts read, ads appear in Pi Browser apps and in App Studio, not in the mining app, and none of them states what funds the company. The MiCA paper’s financial section covers only PiBit, which it says was capitalised by the Pi Foundation.

Why a price is the wrong question

A search for a Pi price forecast lands here; the question under it has no price-shaped answer. A price is what the last trade cleared at, and a trade clears on three assumptions: how much can be sold, who can move the units, and what the holder is owed. For a reserve-backed stablecoin the peg depends on a credible redemption right. For Pi the nearest figure to a float is a circulating supply that counts locked units, with the allocations behind it unlabelled; custody is the project’s record until migration and, by Pi’s account, a non-custodial wallet afterwards; and the MiCA paper states the holder’s claim against the issuer as none. A price on a venue is then the figure traders will pay for a token with no such claim, and it moves with what they believe. MiCA bars a white paper from asserting future value (Article 6(4)), and the sources read here do not supply the inputs a forecast would need.

What to check instead, in the order that changes the answer:

1. Which balance is it? An entry on the ledger and a number in an app are different things, and the first-wallet rehearsal, a small amount sent in and back out, tests a balance by moving it.
2. What does “circulating” count, and who produced the figure? On the explorer the row counts locked units and comes from an off-ledger feed.
3. Who validates, and whom do they trust? A number, names and a trust configuration would answer it.
4. Where are the allocations? Addresses and locks the ledger shows, not proportions in a document.
5. What is the holder owed? The MiCA paper’s answer is nothing from the issuer.
6. What date is the claim? Pi’s pages disagree about which phase is current, so the dated posts and the ledger decide.

Nothing here is a view on what Pi is worth. It marks which claims about it can be checked today and which rest on the project’s word.