NEAR Protocol's upcoming network upgrade 2.13, scheduled for June 2026, bundles two technically distinct capabilities into a single release: dynamic resharding and post-quantum cryptographic signing. The first automates the network's ability to scale horizontally without governance votes or manua...
"When a shard hits a state size threshold, it splits deterministically and is validated by state witnesses with no human intervention." — Bowen Wang, NEAR Protocol Core Developer
NEAR Protocol's upcoming network upgrade 2.13, scheduled for June 2026, bundles two technically distinct capabilities into a single release: dynamic resharding and post-quantum cryptographic signing. The first automates the network's ability to scale horizontally without governance votes or manual validator coordination. The second makes NEAR the first major layer-1 to ship NIST-standardized post-quantum signatures to production.
The upgrade arrives at a moment of renewed attention for NEAR. The token surged 54% in a single week in late May after BitMEX co-founder Arthur Hayes named it part of his "holy trinity" alongside Hyperliquid and Zcash. Grayscale filed an S-1 in January 2026 to convert its NEAR Trust into a spot ETF (ticker: GSNR), currently pending SEC review. The Bitwise NEAR Staking ETP in Europe has grown to roughly $40 million in AUM, absorbing $7 million in inflows in one week alone. Yet the network's TVL remains approximately $160 million — a fraction of its $500 million peak — and market capitalization has contracted from $11 billion to roughly $3.4 billion.
The question is whether infrastructure upgrades can close the gap between technical capability and economic activity.
Until upgrade 2.13, adding shards to NEAR required validators to coordinate manually, hold governance votes, and wait through multi-stage approval processes. The mechanism worked but introduced latency between demand spikes and capacity responses.
Dynamic resharding eliminates this bottleneck. The protocol monitors the state size of each shard in real time. When a shard crosses a predetermined capacity threshold, it splits deterministically — no human coordination, no prolonged voting periods. New shards are validated by state witnesses automatically. The mechanism also works in reverse: underutilized shards can merge, reducing overhead during periods of low activity.
According to NEAR's technical documentation, the architecture can scale to more than 70 shards, with throughput that the team claims would exceed Visa's processing capacity. The practical ceiling depends on validator economics and cross-shard communication latency, both of which remain untested at that scale.
The upgrade ships as part of network version 2.13, which also introduces post-quantum-safe signing. Both features were developed in parallel but serve distinct objectives: one addresses horizontal throughput, the other addresses long-term cryptographic security.
Dynamic resharding represents the culmination of NEAR's multi-year Nightshade sharding roadmap, which has progressed through several phases:
Phase 1 introduced Chunk-Only Producers — validators that could process individual shard chunks without maintaining state for the entire network. This lowered hardware requirements for participation.
Nightshade 2.0 launched stateless validation on mainnet, the largest protocol change since NEAR's October 2020 genesis. Validators no longer need to maintain the full state of a shard locally. Instead, they retrieve necessary state data directly from the network for each block. According to the NEAR team, this reduced validator hardware requirements substantially and improved decentralization metrics.
Phase 3 — Dynamic Resharding (upgrade 2.13) completes the original sharding vision. The network becomes self-adjusting: capacity expands and contracts based on actual utilization rather than governance predictions about future demand.
The practical significance is economic. Fixed-shard architectures force networks to choose between over-provisioning (validators maintain excess capacity at cost) and under-provisioning (congestion during demand spikes). Dynamic resharding shifts this from a governance problem to an algorithmic one.
The same upgrade introduces post-quantum-safe signing using FIPS-204 (ML-DSA, formerly CRYSTALS-Dilithium), a lattice-based digital signature algorithm standardized by NIST in August 2024.
NEAR's account model provides a structural advantage here. Unlike Bitcoin and Ethereum, where wallet addresses are cryptographically tied to elliptic-curve keypairs, NEAR accounts are abstracted from their underlying key material. This means NEAR can add new signature schemes — including post-quantum algorithms — without forcing account migration or breaking backward compatibility.
The Defuse team, which operates NEAR's cross-chain swap infrastructure (NEAR Intents), is separately working to extend quantum-safe Chain Signatures to all Intents users regardless of originating chain. If completed, this would provide post-quantum protection for cross-chain transactions routed through NEAR's infrastructure.
The harder engineering challenge lies downstream. Wallets, APIs, hardware devices, and user workflows must all accommodate significantly larger key and signature sizes inherent to lattice-based cryptography. ML-DSA public keys are approximately 1,952 bytes and signatures approximately 3,293 bytes — compared to 33 bytes and 64 bytes for ECDSA. This is not a trivial integration burden.
NEAR's post-quantum push coincides with an acceleration in the quantum threat timeline. In March 2026, Google Quantum AI published research estimating that breaking 256-bit elliptic-curve cryptography could require roughly 20 times fewer physical resources than previously believed. Where experts once projected a need for tens of millions of qubits, Google's revised estimate is fewer than 500,000 physical qubits under standard error-correction assumptions.
According to Project Eleven's 2026 quantum threat report, the baseline estimate for "Q-Day" — the moment a cryptographically relevant quantum computer can break existing public-key cryptography — sits at 2033, with optimistic and pessimistic bounds of 2030 and 2042 respectively.
Blockchain systems face particular exposure. Addresses hold meaningful value under the same public key for years. Once a key is compromised, there is no recovery mechanism. The "harvest now, decrypt later" attack vector — where adversaries record encrypted data today for future decryption — is especially relevant for high-value blockchain addresses.
NEAR's early adoption of FIPS-204 does not eliminate quantum risk, but it positions the protocol ahead of most layer-1 competitors on migration timelines. Ethereum's post-quantum research (pq.ethereum.org) remains in the proposal stage. Bitcoin has no formal post-quantum migration path.
In May 2026, NEAR Foundation launched four products that pivot the network toward AI-agent commerce:
NEAR AI also introduced automatic PII anonymization for AI prompts, scrubbing personally identifiable information before it reaches inference infrastructure. The confidential GPU marketplace addresses a specific enterprise bottleneck: most institutional AI workloads cannot run on permissionless infrastructure because the data is regulated. TEE-based computation offers a potential bridge.
The revenue thesis hinges on NEAR Intents, the protocol's cross-chain intent-resolution layer. Every swap across more than 35 connected networks routes fees back to the protocol. On February 23, 2026, the fee conversion mechanism activated, routing all Intents fees into NEAR token purchases. To date, Intents has directed revenue toward 2.1 million NEAR in buybacks and burns.
Whether AI agent commerce generates sufficient volume to matter remains an open question. The trailing 90-day average for daily Intents volume through February 2026 was approximately $77 million — short of the $177 million daily threshold needed for net deflation at current prices.
NEAR's tokenomics underwent a structural shift in late 2025. On October 30, 2025, the inflation rate was permanently reduced from 5% to 2.5%, cutting annual issuance roughly in half to approximately 32 million tokens per year.
The deflationary model depends on three variables: Intents volume, fee rates, and token price. At the current channel-mix-weighted fee rate, the deflationary threshold sits at approximately $177 million in daily Intents volume. The trailing 90-day average through February 2026 was $77 million — 56% below the breakeven point.
Growth projections from SVRN Research suggest daily volume could reach $100–150 million by end of 2026 under a conservative 50–100% compound annual growth rate. Under this model, NEAR could cross into net deflationary territory by 2027. The model's assumptions about sustained growth rates remain untested.
Current market metrics:
NEAR occupies a specific niche among sharded layer-1 architectures:
| Feature | NEAR | Ethereum | Polkadot | Cosmos | |---------|------|----------|----------|--------| | Sharding model | Dynamic (auto-split/merge) | Rollup-centric (danksharding planned) | Parachains (fixed slots) | App-chains via IBC | | Shard count | Adaptive (target: 70+) | N/A (rollup-based) | ~50 parachains | 100+ IBC chains | | Post-quantum status | FIPS-204 shipping June 2026 | Research stage (pq.ethereum.org) | No formal plan | No formal plan | | Validator requirements | Stateless (low hardware) | Full state (high hardware) | Shared security | Sovereign sets |
The Grayscale NEAR Trust ETF conversion (S-1 filed January 20, 2026, ticker GSNR) is pending SEC review. If approved, it would list on NYSE Arca and provide regulated U.S. exposure to NEAR — a potential catalyst for institutional inflows. The Bitwise NEAR Staking ETP already holds ~$40 million AUM in Europe.
Arthur Hayes' May 22 "holy trinity" endorsement (NEAR, HYPE, ZEC) preceded the 54% weekly surge. However, on June 4, Hayes disclosed he had sold his entire NEAR position, citing macro concerns including rising energy prices and anticipated AI IPO-related market volatility.
NEAR's upgrade 2.13 is technically substantive. Dynamic resharding completes a multi-year sharding roadmap that began with Nightshade Phase 1 and progressed through stateless validation in Nightshade 2.0. Post-quantum signing addresses a real cryptographic timeline that Google's 2026 research made more urgent.
The gap between infrastructure capability and economic traction persists. A network that can auto-scale to 70+ shards currently holds $160 million in TVL. A protocol with 46 million monthly active users generates $77 million in daily Intents volume against a $177 million deflationary threshold. The AI commerce layer — Agent Market, IronClaw, confidential GPU marketplace — is architecturally coherent but pre-revenue in any meaningful sense.
The economic value question is straightforward: can NEAR convert technical differentiation into fee revenue at scale? The infrastructure is being built. The demand has not yet materialized at the levels required to sustain the tokenomic model. The data will resolve this within 12–18 months.