BNB Chain

BNB Chain traces its history to Binance Chain, whose mainnet launched on April 18, 2019. Designed for asset issuance and decentralized trading, it used Tendermint BFT and delegated proof of stake, targeting high-throughput order matching and transfers. BNB, originally an Ethereum ERC-20 token, migrated to the chain as a native BEP-2 asset. Binance Chain’s limited scripting and smart-contract capabilities later helped motivate a parallel EVM-compatible chain, Binance Smart Chain.⁠[2]

Binance Smart Chain (BSC) launched on September 1, 2020 alongside, rather than as a direct replacement for, Binance Chain. BSC provided EVM compatibility and smart-contract execution, enabling developers to migrate Ethereum tools and applications; BNB paid gas fees and supported validator staking. BSC initially used PoSA with an approximately three-second block interval, trading a smaller validator set for faster blocks. This date marks the start of the current BSC network’s life.⁠[3]

As the BSC ecosystem expanded rapidly in the first half of 2021, it experienced a wave of dApp security incidents. A contemporary rekt.news ranking estimated roughly $220 million in losses across ten incidents involving BSC projects. Uranium Finance lost an estimated $57.2 million after a missing zero in modified Uniswap code allowed an attacker to withdraw disproportionate liquidity; PancakeBunny’s LP-price-dependent rewards were manipulated by flash loan, allowing the attacker to mint and sell BUNNY for an estimated $45 million; Meerkat Finance’s vault initialization permissions let an attacker obtain ownership and withdraw about $32 million; and a pool-share calculation flaw cost Spartan Protocol approximately $30.5 million.⁠[4]

The same reviews covered BurgerSwap reentrancy and repeated exchanges after removal of a constant-product check, Belt Finance strategy-pool balance calculations and price manipulation, another calculation error at Merlin Labs after an initial fix, and repeated attacks on Value DeFi. The recurring patterns included mistakes in forked code, price-oracle dependencies, share or reward calculations, permission initialization and insufficiently audited emergency fixes.⁠[4]

These figures are contemporaneous public estimates and do not consistently deduct recoveries or compensation. Most vulnerabilities lay in applications deployed on BSC, not proof that its consensus layer or underlying protocol was breached. The incidents illustrate the application-layer security costs of early ecosystem expansion and the context for improvements in oracles, audits and incident response.⁠[4]

BSC completed the Bruno upgrade and deployed BEP-95 in December 2021. The proposal burns a governance-set share of each block’s gas fees, providing a real-time burn mechanism separate from quarterly and later Auto-Burn programs. Burned BNB is permanently removed from circulation, while on-chain governance can adjust the share. The mechanism links network use to BNB supply changes, but does not burn every transaction fee; validators receive the remainder.⁠[5]

On February 15, 2022, the ecosystem renamed Binance Chain as BNB Beacon Chain and Binance Smart Chain as BNB Smart Chain, adopting BNB Chain for the overall brand. It also explained BNB as “Build and Build” and emphasized a move toward community-driven infrastructure. The rebranding did not create a new blockchain or change existing BSC addresses and transaction history. Older references to BSC or Binance Smart Chain generally refer to today’s BNB Smart Chain.⁠[6]

The 2022 roadmap proposed BNB Application Sidechains (BAS) for horizontal scaling. BAS used PoSA and EVM compatibility to provide dedicated block space for individual high-usage applications. Projects could customize gas tokens, fees and validator sets and transfer BEP-20 and BEP-121 assets through a native bridge to BSC. BNB Chain’s core team pursued the design with NodeReal and Ankr, targeting applications such as GameFi and SocialFi that could congest shared mainnet space.⁠[7]

The roadmap also proposed raising the mainnet block gas limit gradually to 200 million; BEP-127 validator maintenance mode; BEP-131’s expansion from 21 active validators to a 41-seat structure including candidates; open-sourcing the BNB Beacon Chain DEX matching engine; and faster finality through BEP-126 BLS multisignatures.⁠[7]

A later phase called zkBAS envisioned BEP-20 aggregation and NFT operations first, then general ZK circuits for specific applications, a target throughput approaching 10,000 transactions per second and research into inexpensive non-ledger storage. These capacities, throughput figures and timelines were plans and design goals in 2022, not evidence that all features launched or reached the stated performance.⁠[7]

On October 6, 2022, an attacker exploited a proof-verification flaw in BSC Token Hub, the bridge between BNB Beacon Chain and BSC, to mint approximately 2 million BNB improperly. Validators coordinated a pause in BSC block production to limit further transfers, and the network resumed after patches and validator upgrades. An official ecosystem update estimated that $100 million to $110 million in assets had moved off-chain. The incident showed systemic risk in bridge verification and raised questions about decentralization, emergency governance and censorship when validators can coordinate a halt.⁠[8]

The OP Stack-based opBNB mainnet launched in September 2023 as a BSC layer 2 for higher-frequency transactions and applications. BNB Greenfield launched in October for decentralized storage, access control and a data economy. Together with BSC they formed the “One BNB” multichain framework: BSC supplies layer-1 settlement and smart contracts, opBNB layer-2 execution capacity, and Greenfield programmable storage. BNB Chain thus became an umbrella for computing, scaling and storage components.⁠[9]

Beginning in May 2024, BNB Chain pursued “BNB Chain Fusion” to move staking, governance and token functions from BNB Beacon Chain to BSC, ending the dual-chain architecture and unifying the technology stack. Native staking, governance and cross-chain assets required migration, and users and projects were asked to convert BEP-2 and BEP-8 assets still on Beacon Chain before its retirement. Fusion reduced the complexity of maintaining and bridging two layer-1 chains, while recovery of legacy assets required separate tools and deadlines.⁠[10]

BSC implemented BEP-322 in May 2024, introducing proposer-builder separation (PBS) and an independent builder market. Unlike Ethereum’s common relay path, BSC used validator-run mev-sentry services to interact directly with builders because of its smaller validator set and shorter intervals. Validators could inspect bid blocks, and each builder could send at most three bids per interval. BSC’s implementation barred builders from changing the coinbase address to their own and allowed zero-gas-price transactions subject to a minimum effective block gas price.⁠[11]

BlockSec reported in January 2025 that eight builders participated, with Blockrazor and 48Club together building over 80% of blocks and Bloxroute, Blocksmith and NodeReal roughly 19%. A BSCScan snapshot showed 45 validators, including 21 Cabinet and 24 Candidate validators, staking about 29.24 million BNB. In a December 1–18, 2024 sample, the authors also observed close ties between some builders and validators, suggesting concentration of transaction flow and vertical-integration risk.⁠[11]

PBS can allow users to send transactions to private pools visible only to builders, reducing public mempool exposure, but it does not automatically protect users who remain in public pools. Citing Dune, BlockSec said BSC sandwich-attack transaction counts rose after May 2024 and attributed part of the increase to bundles lowering the cost of failed attacks; it also recorded phishing-contract abuse of a zero-Gwei channel. Builder shares, validator counts and attack trends are observations from particular periods, not permanent structural facts or proof that BEP-322 caused every risk.⁠[11]

On June 20, 2024, BSC’s Haber upgrade introduced BEP-336 blob transactions. Drawing on Ethereum’s EIP-4844, it places batch data submitted by layer-2 networks in temporary blobs rather than permanently in ordinary block data, reducing data-availability costs and improving the fee structure for networks such as opBNB. Official pre-launch cost-reduction figures were design estimates and test expectations, not proof of equal mainnet savings for every application and period.⁠[12]

As Fusion concluded, BNB Beacon Chain stopped producing blocks on December 3, 2024 and its principal functions were consolidated on BSC. The independent consensus chain from the Binance Chain era no longer produced new blocks, although historical data remained accessible. Cross-chain pegged tokens not migrated in time do not automatically recover and require a dedicated recovery process, a distinction between retiring a chain and ending support for ordinary software.⁠[13]

The Pascal hard fork activated on BSC mainnet at 02:10 UTC on March 20, 2025. It incorporated EIP-7702, allowing externally owned accounts to temporarily set contract code in a transaction, providing a basis for batch calls, sponsored gas and account-abstraction experiences. It also added Ethereum-compatibility changes including BLS12-381 precompiles. Greater wallet programmability makes clear authorization displays important because malicious authorization can expand account risk.⁠[14]

The Lorentz hard fork activated at 05:05 UTC on April 29, 2025. BEP-520 cut the target block interval from about three seconds to 1.5 seconds, alongside validator-network and consensus improvements. Faster blocks can reduce the wait for transaction inclusion but leave validators less time to propagate blocks and execute transactions, making node performance, bandwidth and consensus stability ongoing concerns.⁠[15]

On June 30, 2025, the Maxwell hard fork further reduced BSC’s target block interval from 1.5 to 0.75 seconds. BEP-524 adjusted the interval, while BEP-563 and BEP-564 improved validator communications and node synchronization to ease the pressure of subsecond production. Confirmation and finality waits could theoretically fall further, but official performance figures included test results and targets rather than a sustained mainnet throughput guarantee.⁠[16]

As of December 8, 2025, a report using Token Terminal and DeFiLlama data put BNB Chain daily active addresses at about 3.7 million, with a yearly peak near 5 million, and opBNB daily active addresses at about 1.5 million. Cumulative DEX volume exceeded $2 trillion; 2025 volume through December 8 was about $680 billion and October volume about $119.2 billion. These third-party figures can vary with deduplication, chain coverage and subsequent revisions.⁠[17]

Artemis and Visa Onchain Analytics figures put combined BSC and opBNB stablecoin supply at about $14 billion on December 6, up 102.9% from roughly $6.9 billion at the start of the year. Year-to-date stablecoin volume was approximately $953.8 billion and monthly volume averaged about $85 billion from January through November. The report linked growth to a “0 Gas Fee Carnival” for assets including USDT, USDC and USD1, low fees and exchange ecosystem access; that association alone does not establish causation.⁠[17]

An RWA.xyz snapshot put on-network real-world asset market value at about $1.74 billion on December 8, across more than 100 projects and approximately 5,231 holders. Launches included over 60 tokenized stocks and ETFs through xStocks in July, over 100 assets through Ondo Global Markets beginning in October, BlackRockBlackRockBlackRock is a publicly traded investment management company headquartered in New York, founded in 1988 by Larry Fink and seven others and listed on the New York Stock Exchange (ticker BLK); its businesses include the iShares family of exchange-traded funds and the Aladdin investment and risk management platform. As of June 30, 2026, its assets under management totaled $15.3 trillion.Open the full entry BUIDL via Securitize and Wormhole in November, and China Merchants Bank International money-market fund tokens and Circle USYC. The network also established an initial $50 million RWA incentive program. These year-end snapshots do not imply risk-free assets, uniform global compliance, or unchanged market values and liquidity.⁠[17]

BNB Chain said it was preparing a new “native stablecoin” intended to connect liquidity across DeFi, payments, settlement and gaming and aimed at large-scale applications. Its announcement said only “coming soon” and did not disclose a name, peg, legal issuer, reserve composition, custodian banks, redemption mechanism, auditor or launch date.⁠[18]

The announcement therefore recorded a proposed product and liquidity direction, not an issued or licensed stablecoin or verifiable reserves. It was distinct from third-party stablecoins such as USDT, USDC and USD1 already circulating on BNB Chain.⁠[18]

The Fermi hard fork activated on BSC mainnet at 02:30 UTC on January 14, 2026. BEP-590 reduced the target interval from 0.75 to 0.45 seconds; other proposals adjusted fast finality, consensus voting and validator synchronization. In less than a year BSC had moved from three seconds to 1.5, 0.75 and then 0.45 seconds. Faster production does not make every transaction immediately irreversible: inclusion, sufficient confirmations and protocol finality are distinct.⁠[19]

On March 4, 2026, BNB Chain announced a retirement plan for its Beacon Chain Token Recovery Tool, which helped eligible users recover pegged tokens not migrated before Beacon Chain retired. The plan set phased deadlines and said recovery services would eventually stop. It marked a closing phase for Fusion’s legacy-asset processing and reminded holders of old BEP-2, BEP-8 and cross-chain pegged assets that recovery is not permanent; eligibility, contract addresses and official deadlines require verification.⁠[20]

The Osaka/Mendel hard fork activated on BSC mainnet at 02:30 UTC on April 28, 2026. Rather than shorten the 0.45-second interval again, it focused on execution consistency, gas metering, finality and Ethereum compatibility. Of nine BEPs, six came from Ethereum EIPs alongside BNB Chain-specific changes. The upgrade reinforced the importance of stable, predictable execution and node synchronization after subsecond acceleration.⁠[21]

BNB Chain published the design and test results of a post-quantum cryptography migration prototype for BSC. It replaced secp256k1 ECDSA transaction signatures with NIST FIPS 204 ML-DSA-44 and BLS12-381 aggregate signatures in fast-finality votes with aggregated pqSTARK proofs. ML-DSA-44 public keys are 1,312 bytes and signatures 2,420 bytes; a new 0x05 transaction type used an on-chain 0x70 public-key registry. Addresses remained 20 bytes, derived from the ML-DSA public key using Keccak-256, so the report said RPC, SDKs and address length need not change with the prototype.⁠[22]

Size was the main tradeoff. The report said a native transfer grew from about 110 bytes to 2.5 KB and a block from around 130 KB to 2 MB at 2,000 native transfers per second. In cross-region testing, peak native-transfer throughput fell from a non-post-quantum baseline of 4,973 TPS to 2,997 TPS, about 40%, while gas processed per second fell from 392 to 196, about 50%; mixed-load throughput fell from 3,695 to 2,406 TPS, about 35%. Roughly 14.5 KB of raw signatures from six validators could be aggregated into a 340-byte proof, a compression ratio of about 43:1. ConsensusConsensusConsensus is an annual cryptocurrency, blockchain, and Web3 industry conference hosted by the crypto media outlet CoinDesk (not the consensus mechanism of a blockchain). It was first held in New York in 2015 and later moved to Austin and Toronto; a Hong Kong edition was added from 2025, and since 2026 the main conference has been held in Miami.Open the full entry-proof size was manageable, but transaction signatures remained a networking and data bottleneck.⁠[22]

The prototype covered only transaction signatures and consensus votes. P2P handshakes still used secp256k1 ECDH, with ML-KEM a planned direction; EIP-4844 KZG commitments still depended on BLS12-381 pairings, with lattice-based commitments a future direction. Both were outside the reported code scope. The related bsc pull request #3660 was closed, not merged and released. The report concluded that current approaches could achieve post-quantum readiness but network and data-layer scaling remained major production challenges. The prototype and tests did not mean that post-quantum transactions or consensus had been enabled on BSC mainnet.⁠[22]

On July 8, 2026, BNB Chain published its second-half technical roadmap. It described 0.45-second blocks, fast finality and benchmark throughput as first-half milestones and planned a next-generation architecture testnet for late 2026 and a mainnet target in early 2027. It also listed post-quantum cryptography research into long-term risks to existing signatures. This was research and planning, not deployed quantum-safe signatures; test benchmarks are not sustained mainnet measurements.⁠[23]

The roadmap described next-generation L1 goals of over 100,000 transactions per second through combined optimization of consensus, parallel execution and LtHash-based storage; preconfirmation below 50 milliseconds; and block finality below one second. TxStream proposed removing the public mempool and streaming transactions directly to the block leader to lower latency and prevent front-running by design. PriorityLane proposed reserving block space through on-chain governance for critical traffic such as oracles, liquidations and bridges.⁠[23]

It also listed protocol-level confidential transactions and selective disclosure, sponsored gas, GasToken, transaction batching, scheduled execution, passkey signatures and access-key controls, saying the new architecture would remain BNB-powered and interoperable with the existing ecosystem. The throughput, latency and privacy figures were design goals, with a testnet planned for late 2026 and mainnet targeted for early 2027; they were not measured sustained capabilities of BSC mainnet in July 2026.⁠[23]

BNB Chain said a former employee created a wallet and issued the TST meme coin for a video tutorial, then left the organization. According to its statement, the former employee retained the seed phrase and could generate new private keys without authorization; the same address was subsequently used independently in activities involving the new BNB Chain meme coin ASTEROID.⁠[24]

BNB Chain said it did not create, authorize, promote or participate in ASTEROID’s issuance and did not control the token or wallet. It said it was taking legal action against the former employee and cooperating with relevant organizations. The identity, source of access and legal action are the network’s account, not a judicial finding of liability, recovery or criminal conviction. The episode also raised key-lifecycle and access-revocation risks when a departing employee retained a tutorial wallet’s seed phrase.⁠[24]

On-chain monitor Lookonchain said four newly created wallets bought 796.7 million ASTEROID, or 79.67% of supply, for roughly $10,000, then sold 718.8 million for 1,103 BNB, valued at about $638,000 at the observation time, for an estimated $628,000 profit.⁠[25]

The figures describe purchase concentration, sales and the return path of BNB at public addresses. The identity link, cost attribution and profit calculation combine the monitor’s estimates and the official statement; they are not a judicial determination of illegal proceeds. Dollar values change with BNB’s price.⁠[25]

The BNB Chain BEP repository contained draft BEP-707, “Post-Quantum Cryptography Migration Roadmap for BNB Chain,” an informational umbrella proposal dividing migration into independently reviewable and activatable BEPs. Stage one upgrades P2P key exchange; stage two progressively binds post-quantum keys to existing ECDSA accounts while retaining addresses; stage three migrates BLS aggregation, KZG commitments and BN254 pairing precompiles. The ordering reflects accumulated retrospective risk and maturity, not difficulty alone: P2P is exposed to “collect now, decrypt later” risk without touching consensus; account signatures affect assets but allow gradual migration; pairing cryptography lacks mature post-quantum aggregation replacements and has no fixed activation date.⁠[26]

Affected components include RLPx ECDH key exchange, secp256k1 ECDSA for accounts and transactions, BLS12-381 aggregate signatures for fast finality, KZG commitments for layer 2 and rollups, and BN254 pairing precompiles 0x06–0x08. The draft distinguishes the reduction in effective hash security under Grover’s algorithm for Keccak and SHA-2/3 from Shor’s effects on ECDSA, ECDH, BLS and pairing structures.⁠[26]

A post-quantum core chain would not automatically protect applications. Bridges, stablecoin and RWA issuers, DAO treasuries, multisig wallets, oracles and contracts verifying signatures or pairing proofs must migrate separately; non-upgradeable contracts hard-coded to old algorithms cannot be repaired solely by a chain upgrade. Each stage envisaged a hybrid transition with classical and post-quantum cryptography, and governance for changes affecting consensus, economics or compatibility. At the entry’s update, the BEP-707 pull request was open and its status Draft, not an approved hard-fork plan or completed deployment.⁠[26]

Draft BEP-708 detailed the first BEP-707 stage with an RLPx v5 handshake combining X25519 ECDH and NIST FIPS 203 ML-KEM-768. Classical, X25519 and ML-KEM shared secrets feed HKDF-SHA256, binding the full handshake transcript before deriving session keys. The design adds kem-group and kem-share to auth and ack messages, using EIP-8’s rule allowing older clients to ignore trailing fields for compatibility. Failed handshakes must not automatically trigger downgrade; a pqkex flag in signed ENRs and pq-capable in encrypted Hello messages detect abnormal downgrades.⁠[27]

Because the proposal does not change blocks, transactions, state, gas, EVM or consensus rules, it envisaged phased client releases without a hard fork. Stage 0 is testnet-only and off by default; stage 1 prefers v5 and measures mainnet post-quantum sessions; stage 2 disconnects detected downgrades only after at least 90% mainnet adoption for four consecutive weeks; stage 3 plans strict default v4 rejection only after 99% adoption and at least six months of stage 2. The draft estimates a full handshake growing from about 0.70–0.90 KB to 3.05–3.25 KB, an added 2.3 KB per TCP connection rather than per P2P message.⁠[27]

This stage protects RLPx session contents against “record now, decrypt later” attacks, while node identity authentication remains secp256k1. An attacker with a usable quantum computer could still impersonate nodes in real time; account and node identity signatures are left to stage two. The pull request remained open and Draft, without a merged implementation, client release date or mainnet adoption rate. A compatibility design and four-stage thresholds did not mean the P2P network was already quantum-safe.⁠[27]

The Pasteur hard fork entered its BSC mainnet upgrade window at 02:30 UTC on August 25, 2026. BEP-682 and BEP-695 strengthened bridge, validator, staking and governance paths, while BEP-675 increased execution capacity for full blocks. QANet throughput changes announced before launch were test-network results, not long-term mainnet guarantees. Pasteur continued the 2026 shift from further block-time reductions toward security, governance and capacity on a subsecond foundation.⁠[28]

New L1 documentation described an EVM-compatible chain alongside BSC using BNB without issuing a new token. Its targets were 200-millisecond blocks, approximately 20-millisecond preconfirmations and subsecond ordering finality; execution finality is distinct. A multi-validator devnet was running, while public testnet, mainnet and the native bridge were not yet ready.⁠[29]