NEAR Protocol's v2.13 network upgrade, scheduled for June 2026, introduces two capabilities that address distinct infrastructure gaps: dynamic resharding for automatic horizontal scaling and FIPS-204 (ML-DSA) post-quantum cryptographic signing. The upgrade represents the final phase of a five-yea...
"The users of blockchain will be AI agents, not humans." — Illia Polosukhin, Co-founder, NEAR Protocol (CoinDesk, March 2026)
NEAR Protocol's v2.13 network upgrade, scheduled for June 2026, introduces two capabilities that address distinct infrastructure gaps: dynamic resharding for automatic horizontal scaling and FIPS-204 (ML-DSA) post-quantum cryptographic signing. The upgrade represents the final phase of a five-year sharding roadmap that began with Simple Nightshade in 2021 and progressed through stateless validation in 2024.
The token has responded. NEAR traded at approximately $2.13 on June 14, with a market capitalization of $2.77 billion — up 262% from its 2026 low. NEAR Intents, the protocol's cross-chain execution layer, has processed over $20 billion in cumulative transaction volume and $32 million in fees. Whether the technical upgrades justify the price action depends on whether dynamic resharding generates measurable demand or remains an unused capacity expansion.
Dynamic resharding allows NEAR's network to automatically add or remove shards based on real-time demand without governance votes, manual coordination, or protocol upgrades.
The mechanism operates deterministically: when a shard's state size exceeds a defined threshold, it splits. State witnesses validate the new shard configuration. No human intervention is required.
Previously, adding shards required a full protocol upgrade involving weeks of validator coordination, a governance vote, and a staged rollout. The network currently runs nine live shards on mainnet, up from six in 2024 and eight in early 2025. Each shard addition before v2.13 was a discrete engineering and governance event.
Dynamic resharding eliminates this bottleneck. If traffic spikes — whether from human users, DeFi protocols, or autonomous agents — the network absorbs the load by splitting overloaded shards automatically. Conversely, underutilized shards can be merged to reduce validator overhead.
The practical implication: NEAR no longer needs to predict demand. The network provisions capacity in response to usage, similar to how cloud infrastructure auto-scales compute instances.
NEAR's sharding architecture has evolved through four distinct phases:
| Phase | Date | Capability | |-------|------|-----------| | Simple Nightshade (Phase 0) | September 2021 | Initial shard architecture with 4 shards | | Phase 1 | 2022 | Chunk-only validators; 86% reduction in collateral requirements | | Nightshade 2.0 (Phase 2) | January 2024 | Stateless validation; lowered hardware requirements | | v2.13 (Dynamic) | June 2026 | Automatic shard splitting and merging based on demand |
The progression follows a clear pattern: each phase reduced the cost or complexity of running the network. Phase 1 lowered the capital barrier to validation. Phase 2 lowered the hardware barrier. Phase 3 removes the coordination barrier entirely.
This trajectory positions NEAR as one of the few Layer-1 protocols to have shipped a multi-year sharding roadmap to completion. Ethereum's danksharding remains in development. Zilliqa, an early sharding proponent, has not implemented dynamic resharding. Polkadot uses parachains rather than homogeneous shards.
The same v2.13 upgrade introduces post-quantum cryptographic signing based on FIPS-204 (ML-DSA, formerly CRYSTALS-Dilithium), a lattice-based digital signature algorithm standardized by NIST in August 2024.
NEAR's implementation allows users to upgrade their account security to quantum-resistant signing with a single transaction. This is possible because NEAR accounts use human-readable names and rotatable access keys, unlike Bitcoin and Ethereum addresses where changing the signing scheme requires migrating to a new address.
The quantum threat to blockchain is not imminent. NIST's draft guidance in IR 8547 recommends deprecating quantum-vulnerable algorithms (RSA, ECDSA, EdDSA) by 2030 and fully retiring them by 2035. No blockchain network faces an active quantum attack vector today.
However, a specific risk applies to blockchain: the "harvest now, decrypt later" threat. An adversary could record encrypted transactions today and decrypt them when quantum computing matures. For long-duration custody — assets held for years or decades — the threat window is relevant.
NEAR is not the first blockchain to address this. QANplatform and others have explored post-quantum signatures. But NEAR appears to be the first top-30 Layer-1 by market capitalization to ship a NIST-standardized post-quantum signing scheme to production.
NEAR Intents, the protocol's cross-chain execution layer, crossed $20 billion in cumulative transaction volume in early June 2026. The system facilitates threshold signatures across 35+ chains and has processed over 25 million swaps.
Key metrics as of June 2026:
The second $10 billion in volume was processed in approximately four months, indicating accelerating adoption relative to the first $10 billion. However, the 0.16% take rate is thin. At $32 million in cumulative fees against $20 billion in volume, the economic value captured by the protocol remains modest relative to the infrastructure investment.
For context, daily protocol fees on NEAR totaled approximately $137,150 on a recent trading day. Annualized, that is roughly $50 million — not immaterial for a $2.77 billion market cap protocol, but not sufficient to justify the valuation on fee revenue alone.
NEAR's strategic narrative centers on AI agents as the primary future users of blockchain infrastructure. Co-founder Illia Polosukhin — one of the eight co-authors of the 2017 "Attention Is All You Need" paper that introduced the transformer architecture underlying all modern large language models — has positioned NEAR as the "currency of agents" and the network as a "unified commerce layer."
In May 2026, NEAR rolled out four AI-focused product launches:
NEAR also joined NVIDIA's Inception Program in January 2026, gaining access to GPU resources and developer tooling for privacy-preserving AI systems.
The thesis is logically coherent: if AI agents become the dominant on-chain transactors, they need infrastructure that scales automatically (dynamic resharding), executes across chains (Intents), and preserves data privacy (TEE + MPC stack). NEAR is building all three.
The question is timing and competition. Solana reportedly handles 65% of AI agent payment transactions as of mid-2026. Ethereum's rollup ecosystem offers comparable cross-chain execution through various bridge and intent protocols. NEAR's AI infrastructure is differentiated but unproven at scale.
NEAR traded at approximately $2.13 on June 14, 2026, with a market capitalization of $2.77 billion. The token is up 262% from its 2026 low, driven by:
The Hayes endorsement warrants scrutiny. On May 22, Hayes publicly labeled HYPE, ZEC, and NEAR as his portfolio's "holy trinity" and established long positions. By June 5, he had completely unwound all three positions. Blockchain investigator ZachXBT accused Hayes of using followers as exit liquidity, a claim Hayes has not publicly addressed.
The token's 24-hour trading volume surged to over $1.4 billion during the rally, suggesting substantial speculative interest beyond fundamental demand.
Utilization risk. Dynamic resharding solves a capacity problem NEAR does not yet have. The network runs nine shards. Whether demand materializes to require automatic scaling is uncertain. Building capacity ahead of demand is a legitimate infrastructure strategy, but it does not generate revenue.
TVL decline. Total value locked on NEAR dropped from approximately $500 million at its peak to roughly $100 million as of early 2026. DeFi activity on NEAR has contracted, raising questions about whether the AI agent narrative translates to measurable economic activity.
Competitive pressure. Solana's sub-second finality and established DeFi ecosystem, Ethereum's rollup-centric roadmap, and emerging chains like Monad all compete for the same AI infrastructure use case. Polosukhin's transformer paper credentials provide narrative advantage but not a technical moat.
Fee revenue gap. At $137,150 in daily fees, NEAR's price-to-fee ratio is elevated. The $2.77 billion market cap implies the market is pricing future fee growth that has not yet materialized.
Post-quantum timeline mismatch. NIST's own guidance does not recommend deprecating current algorithms until 2030. NEAR's post-quantum implementation is forward-looking, but the practical security benefit is years away.
NEAR Protocol has shipped what most Layer-1 projects only promise: a working, multi-year sharding roadmap culminating in automatic horizontal scaling. The addition of NIST-standardized post-quantum cryptography adds a forward-looking security layer that no comparable Layer-1 has implemented.
The economic case is less settled. Daily fee revenue of ~$137,000 does not support a $2.77 billion valuation on fundamentals alone. TVL has contracted. The AI agent thesis, while logically sound, competes with Solana's existing market share and Ethereum's ecosystem breadth.
What NEAR has built is infrastructure capacity. What it has not yet demonstrated is infrastructure demand sufficient to justify the market's pricing. The v2.13 upgrade removes a technical ceiling. Whether traffic rises to test that ceiling remains the open question.