The updated paper, now on its third revision at arXiv (2505.14914), was authored by Benjamin Marsh, Steven Landers, and Jayendra Jog. It sets a new finality target of under 250 milliseconds, down from the sub-400ms goal in the original V1 document published May 19, 2025. The throughput target remains 200,000 transactions per second (TPS), equivalent to five gigagas per second. Neither Autobahn consensus nor the Sedna private mempool system has launched on mainnet yet; the whitepaper is a technical specification, not a live product release.

What Changed From V1

The original Giga whitepaper introduced Autobahn, Sei's proposed consensus mechanism. Unlike standard blockchains where a single validator proposes each block, Autobahn assigns every validator its own continuous "lane" for streaming transaction batches. A leader validator periodically takes a snapshot called a "tip cut" to finalize ordering through a two-phase BFT (Byzantine Fault Tolerant) vote. The design is meant to eliminate the single-proposer bottleneck that constrains throughput on chains like Ethereum and Solana.

V2 adds two components that V1 left unresolved. The first is Sedna, a private mempool protocol. Rather than broadcasting full transactions publicly before they are included in a block (the current norm on most chains, which enables front-running and sandwich attacks), Sedna splits each transaction into cryptographic fragments using rateless erasure coding. These fragments are distributed privately across subsets of proposer lanes. No single proposer can reconstruct a full transaction until enough fragments have been committed on-chain. The Sei blog described the guarantee this way: "No proposer will see the full contents of a transaction until ordering is finalized."

The second addition is a formal MEV and censorship-resistance layer. MEV, or maximal extractable value, refers to profits validators and bots extract by reordering, inserting, or censoring user transactions. A Flashbots study, as reported in EarnifyHub's 2026 MEV analysis, estimates that roughly 1.2% of all Ethereum DEX trades are sandwiched, costing users an average of 0.41% per trade. The Sedna mechanism design paper (arXiv 2603.17614) claims the system can reduce MEV costs to 0.04% of transaction value under realistic parameters. That figure is modeled, not yet tested in production, and should be treated as theoretical until Sedna launches on mainnet.

The Sedna paper also identifies a potential weakness in its own design, called the "knife-edge pathology," where withholding a single fragment bundle could delay transaction inclusion. The proposed fix, a reward mechanism called PIVOT-KK, concentrates payouts on the specific bundles that trigger successful decoding, making withholding directly costly for validators.

Foundation: The SIP-3 Migration

Before Giga can be deployed, Sei completed a structural prerequisite. SIP-3, a community-approved governance proposal, was approved by the Sei community in May 2025 and activated on June 1, 2026, a gap of more than a year between approval and activation. It deprecated all Cosmos-native messaging and CosmWasm smart contracts, making Sei a pure EVM chain. The Sei blog called the overall Giga upgrade "the most complex blockchain upgrade since Ethereum's Merge." The transition required USDC.n holders to migrate to native USDC and affected exchanges, custodians, and developers building on the network. The payoff is a simplified architecture compatible with standard Ethereum tooling including Hardhat, Foundry, and Remix.

Token and Network Context

SEI was trading at approximately $0.047 at time of publication, giving it a market cap near $333 million and a fully diluted valuation around $492 million. These figures are approximate and should be verified against live market data before acting on them. Circulating supply stands at roughly 6.08 billion of a 10 billion total. Roughly 112 to 132 million tokens are unlocking monthly through mid-2027, creating persistent sell-side pressure for investors and liquidity providers evaluating the network.

On-chain activity tells a mixed story. Daily active addresses grew 93.5% quarter over quarter, and daily transaction counts rose 87%. Total registered accounts surpass 86 million. TVL, however, has collapsed from roughly $626 million in July 2025 to approximately $41.6 million today, an approximately 93% decline. Yei Finance, the largest lending protocol on Sei, holds more than 60% of remaining DeFi TVL on the chain.

Regional Implications

For traders in South Asia and Sub-Saharan Africa, the practical stakes of Sedna are straightforward. MEV functions as a hidden tax on every DeFi trade. For a user in Lagos or Mumbai executing smaller trades in the $50 to $500 range, losing 0.23 to 0.41% per transaction to sandwich bots is proportionally more damaging than it is for institutional participants. If Sedna functions as specified, a private mempool of this design would remove that cost structurally once live on mainnet.

Sub-250ms finality also matters for remittance applications. India, Pakistan, Nigeria, and Kenya together account for some of the largest global remittance corridors. Current Layer 1 settlement times (roughly 12.8 seconds on Solana, 12 to 15 minutes on Ethereum) make real-time on-chain payments impractical for time-sensitive transfers. Sei Giga's targets would bring settlement latency closer to centralized payment rails. Sei also has a previously announced hardware partnership with Xiaomi to pre-install a native wallet on devices, relevant in markets where Xiaomi holds significant smartphone share.

What Comes Next

Sei's public milestone tracker at giga.seilabs.io, as of its last update on May 27, 2026, shows two major execution upgrades (Ares and Eidos) still in progress, with Autobahn testnet, Autobahn mainnet, Sedna deployment, and a final consensus iteration called Hermes all still ahead. The status of in-progress items may have changed since that update, and readers should consult the tracker directly for the latest figures. Sei Labs has framed 2026 as a progressive rollout rather than a single launch event. The TVL decline is the most visible near-term challenge: user growth and transaction volume are rising, but capital retention is not. Whether the technical case made in V2 translates into developer and liquidity inflows will depend on how that gap closes over the coming months.