Uptime and Quantum Resistance, per x402 Rail

We measure how fast and cheap each x402 rail settles. Two other things matter for a payment rail but cannot be benchmarked: has the chain ever gone down, andis it getting ready for quantum. Here is the documented record for all six rails, each cell sourced. This is a reference, not a FillBench measurement, and we keep that line bright.

Documented and sourced 2026-08-24 (citations at the bottom). Uptime counts full mainnet or block-production halts, not single-validator blips where the network kept running.

Why these two matter for x402

x402 is software paying software, synchronously, for a single request. A chain that haltscannot settle that payment while it is down, so uptime is money: a rail that stalls for hours takes every agent transaction with it. And quantum resistance is the decade-horizon question for anyone custodying value on-chain, because today almost every chain secures accounts with elliptic-curve keys that a large quantum computer could eventually break. Neither is something we can clock, but both are worth knowing before you route real payments over a rail.

Uptime: has it ever gone down?

Full mainnet / block-production halts, best record first. Sourced from public post-mortems and reporting (below). A single number would flatten very different records, so we show the count, the worst duration, and the last incident.
Chain	Major halts	Worst	Last incident
Algorand	0	none	never
XRP	~3	~1 hr	Feb 2025
Stellar	1 to 2	~2 hr	2019
Avalanche	2	~2 to 5 hr	Feb 2024
Base	3	~1 hr	Jun 2026
Solana	7+	~30 hr	Feb 2024

Algorand: Zero downtime since the June 2019 genesis block, roughly 7 years and 3.5B+ transactions with no halt, rollback or reorg. Pure proof-of-stake, designed with no single point of failure.
XRP: Brief halts from node crashes or a quorum dropping below threshold (Nov 2021 ~15 min, Nov 2024 ~10 min, Feb 2025 ~1 hr). Recovered each time with no loss of user funds.
Stellar: A consensus stall on May 15 2019 (67 minutes) plus an earlier 2019 freeze. The 2021 event was a partial SDF-validator outage where the network stayed up. Stellar halts rather than risk a fork.
Avalanche: Block-finalization stalls: March 2023 (a v1.9.12 bug stalled the C and X chains) and February 2024 (no blocks for a couple of hours).
Base: Block-production halts tied to its centralized sequencer: Sept 2023 (~44 min), Aug 2025 (~29 min) and June 2026 (~1 hr). An L2 design trade-off.
Solana: A run of congestion and consensus halts from 2020 to 2024, the worst a ~30 hr stop in Jan 2022. Hardened since 2023 (QUIC, stake-weighted QoS, local fee markets); the last major halt was ~5 hr in Feb 2024.

The point a yes/no column would hide: Algorand is the only rail that hasnever halted, but the rest are not equal. XRP and Basehave had brief, minutes-long blips; Solana once stopped for about a day. Same word ("yes, it has gone down"), very different reality.

Quantum resistance: how ready is it?

Post-quantum status, most ready first. The real bar is protocol-native PQ accounts live on mainnet; app-layer or opt-in and roadmap-only are lesser tiers. Verified 2026-08-24, sourced below.
Chain	Status	What is covered	Where	Next
Algorand	Live, protocol-native	Accounts (opt-in) + State Proofs	Mainnet (v5, Aug 2026)	Consensus PQ targeted 2027
Solana	Live, app-layer	PQ vault (Winternitz, opt-in)	Mainnet (Jan 2025)	Not protocol-native
XRP	Roadmap	ML-DSA in testing	Devnet	Full PQC by 2028
Stellar	Roadmap	Soroban verify (stage 1)	Contracts, 2026	Account signers 2027
Avalanche	Roadmap (proposal)	AIP-QR-001 hybrid sigs	Proposal stage	Not scheduled
Base / Ethereum	Roadmap	Nothing shipped	Research / devnet	~2029 first upgrades

The Algorand standout, stated honestly

On both axes the same chain leads. Algorand has a zero-downtime record since 2019 and is the only x402 rail with protocol-native post-quantum accounts live on mainnet (Falcon-1024, shipped in v5 in August 2026), on top of its Falcon State Proofs from 2022. Among these rails that combination is genuinely unusual.

But we will not overclaim it, because most coverage does. Two caveats matter:

The post-quantum accounts are opt-in. Existing Ed25519 accounts are not automatically migrated, so your funds are only quantum-safe if you actively move to a Falcon account.
Consensus is still classical. Algorand's leader selection (VRF) and block voting remain elliptic-curve based; post-quantum consensus is a 2027 target, not shipped.

So the accurate line is: Algorand is the only rail where you can custody funds in a quantum-resistant account today, and full end-to-end post-quantum security is still on the roadmap. Solana is a real second (a live but app-layer opt-in vault); everyone else is roadmap-only.

How to read this page

Documented, not measured. Roadmaps slip, downtime definitions vary, and opt-in is not the same as default. Treat the cells as sourced facts as of 2026-08-24, and click through to verify. For the numbers we actually clock ourselves, real settlement finality and burst throughput with on-chain proof, see the x402 settlement benchmark and the x402 throughput benchmark.

Common questions

Which x402 blockchain has never gone down?

Algorand. It has run with zero downtime since its June 2019 genesis, roughly seven years and billions of transactions with no full halt, rollback or reorg. Every other rail we track has had at least one block-production halt: XRPL and Base a few brief ones, Stellar and Avalanche a couple, and Solana more than seven (the worst about 30 hours in January 2022, though it has been far more stable since 2023). Uptime matters for payments because a chain that halts cannot settle an x402 payment while it is down.

Which blockchain is quantum resistant today?

As of 2026, Algorand is the only x402 rail with protocol-native post-quantum accounts live on mainnet (Falcon-1024, shipped in v5, August 2026), on top of its Falcon-based State Proofs from 2022. Solana has a live but app-layer option (the Winternitz vault). XRPL, Stellar, Avalanche and Ethereum/Base are all roadmap-only, testing on devnets or still at the proposal stage. Two honest caveats even for Algorand: the post-quantum accounts are opt-in (existing Ed25519 accounts are not automatically migrated), and its consensus layer is still classical until a 2027 target.

Is this measured like your other benchmarks?

No, and we label it clearly. Our settlement and throughput numbers are things we time on-chain ourselves. Uptime history and quantum-resistance status cannot be benchmarked, so every cell here is documented from public post-mortems, official docs and reporting, with sources listed at the bottom. Treat this page as a sourced reference, not a FillBench measurement.

Why not just show a yes or no column for downtime?

Because a yes/no would be misleading. It would put a chain with a single brief incident years ago in the same bucket as one with multiple multi-hour halts, records that differ by orders of magnitude. So we show the count, the worst duration and the last incident instead, which is the honest way to compare reliability.

Sources

Uptime. Solana outage history (Helius); Avalanche Mar 2023 (CoinDesk) and Feb 2024; Stellar May 2019 halt (official); XRPL Feb 2025 (The Block); Base outage history (CoinDesk); Algorand zero-downtime record.
Quantum resistance. Algorand v5 post-quantum accounts and 2027 roadmap; Solana Winternitz vault; Ripple XRPL PQC by 2028; Avalanche AIP-QR-001 proposal; Ethereum PQ roadmap.

Disclosure: Some links on this page are affiliate links. If you subscribe through them we may earn a commission at no extra cost to you. It never changes our measurements or rankings; the numbers here come from automated tests you can reproduce (see our methodology). This is not financial advice, and crypto trading carries risk of loss.