Total Value Locked
$41.2B
+8.4%
30D Δ
01
Executive Overview
A consolidated snapshot of the rollup ecosystem at the close of Q1 2026.
UPDATED
19 MAR 2026
Active Rollups
47
+3
NEW THIS Q
Daily Transactions
14.8M
+22.1%
30D Δ
Avg L1 Cost
0.018 ETH
-12.6%
30D Δ
Quarter at a Glance
The rollup ecosystem expanded materially through Q1 2026 with three new production rollups graduating from testnet, bringing the total active count to 47. Aggregate TVL crossed the $41B threshold for the first time, driven primarily by liquidity migration from L1 venues. Throughput across major optimistic and zk rollups averaged 2,340 TPS, a record high for the cohort.
Per-transaction fees on L2 declined 12.6% quarter-over-quarter as EIP-4844 blob utilization stabilized near 86% of the 6-blob target. Validity-proof rollups continued to outpace fault-proof systems on both throughput and fee compression, though the gap narrowed for the third consecutive quarter as fault-proof systems iterated on compression schemes.
Risk indicators remain green across centralization, sequencer uptime, and bridge inventory; however, prover decentralization for zk systems remains an open structural risk and is flagged in Section 06.
Rollup Distribution
02
Performance Metrics
Throughput, latency, and finality figures aggregated across the active rollup cohort.
SAMPLE
N = 47
Aggregate TPS
2,340
+18.2%
30D Δ
Time to Finality
412 ms
-7.1%
FASTER
Sequencer Uptime
99.984%
+0.04%
30D Δ
Blob Utilization
86.3%
+5.4%
30D Δ
Throughput — 90 Day Aggregate
3,000
2,250
1,500
750
0
D-90
D-75
D-60
D-45
D-30
D-15
D-0
- Aggregate TPS
- 7-day Moving Avg
03
Comparative Matrix
Side-by-side reading of the leading rollups by capacity, cost, and finality.
METHOD
7-DAY MEDIAN
Top 10 Rollups — Comparative Performance
| # | Rollup | Type | TVL | TPS | Avg Fee | Finality | Uptime | 30D Δ |
|---|---|---|---|---|---|---|---|---|
| 01 | Arbitrum One | OP | $14.82B | 412 | $0.012 | 1.2 s | 99.99% | +9.4% |
| 02 | OP Mainnet | OP | $8.91B | 348 | $0.014 | 1.4 s | 99.98% | +6.2% |
| 03 | Base | OP | $6.42B | 390 | $0.011 | 1.3 s | 99.99% | +18.7% |
| 04 | zkSync Era | ZK | $3.18B | 214 | $0.006 | 0.4 s | 99.97% | +12.1% |
| 05 | Linea | ZK | $2.04B | 198 | $0.008 | 0.5 s | 99.96% | +22.5% |
| 06 | Starknet | ZK | $1.86B | 182 | $0.009 | 0.6 s | 99.94% | -3.1% |
| 07 | Scroll | ZK | $1.21B | 160 | $0.010 | 0.7 s | 99.95% | +14.0% |
| 08 | Polygon zkEVM | ZK | $0.94B | 142 | $0.013 | 0.8 s | 99.92% | -1.4% |
| 09 | Mantle | OP | $0.82B | 128 | $0.012 | 1.5 s | 99.93% | +4.6% |
| 10 | Blast | OP | $0.78B | 116 | $0.015 | 1.6 s | 99.91% | -5.8% |
04
Throughput Analysis
Capacity ceiling, sustained median, and observed peaks across rollup architectures.
WINDOW
90 DAYS
Sustained TPS by Rollup
Capacity Headroom
Optimistic rollups continue to demonstrate higher sustained throughput owing to mature compression and minimal proof overhead. Validity (zk) systems are closing the gap rapidly — +34% growth over the 90-day window — with prover hardware and recursive aggregation cited as the dominant unlocks.
Sustained-to-peak ratios suggest meaningful headroom across the cohort: median utilization sits at 43% of observed peak capacity, indicating the network is not yet bottlenecked at the rollup tier.
- ›Compression schemes (BLS, span batches) account for 62% of post-4844 fee compression.
- ›Validity proof aggregation reduced average proof cost 71% YoY.
- ›Sequencer pipelining converged at a 200 ms median slot.
OP Peak TPS
1,184
+11.0%
VS Q4-25
ZK Peak TPS
892
+34.4%
VS Q4-25
Median Utilization
43%
+2.1%
30D Δ
05
Fee Economics
Per-transaction costs, blob utilization, and revenue accrual to L1.
UNIT
USD
Fee Breakdown by Rollup Class
| Class | Median Fee | L1 DA | Proof | Op | QoQ Δ |
|---|---|---|---|---|---|
| Optimistic / blob | $0.013 | $0.008 | — | $0.005 | -14.2% |
| Optimistic / calldata | $0.041 | $0.032 | — | $0.009 | -3.6% |
| Validity / Plonk | $0.009 | $0.004 | $0.003 | $0.002 | -22.8% |
| Validity / Stark | $0.011 | $0.005 | $0.004 | $0.002 | -16.4% |
| Validity / Halo2 | $0.010 | $0.004 | $0.004 | $0.002 | -19.1% |
| Hybrid | $0.018 | $0.011 | $0.004 | $0.003 | -7.7% |
Data Availability Split
L1 Revenue (Q1)
$186.3M
+11.4%
QoQ
L2 Revenue (Q1)
$412.7M
+19.8%
QoQ
L1 Gas Saved
94.2%
+1.2%
30D Δ
L2:L1 Tx Ratio
11.4×
+1.6×
YoY
06
Risk Assessment
Structural, operational, and economic risks observed across the rollup cohort.
SEVERITY
L1 - L4
Sequencer Centralization
Most production rollups still operate a single sequencer. Decentralized sequencer schemes are in production on three rollups; broader rollout expected by Q3.
Prover Decentralization
Validity rollups remain dependent on a small set of prover operators. Open prover networks are emerging but unproven at scale.
Bridge Inventory
Canonical bridges remain healthy with high collateralization ratios; third-party bridges show pockets of low liquidity in long-tail assets.
Aggregate Risk Posture
Aggregate operational risk has trended downward for the third consecutive quarter, with sequencer uptime and bridge collateralization at multi-year highs. The dominant remaining structural concern is prover decentralization for validity systems — a topic on which we expect significant production movement during 2026 as multiple permissionless prover marketplaces approach mainnet readiness.
Economic risk remains contained: L2 fee compression has not materially impacted L1 settlement revenue, and the L2:L1 transaction ratio at 11.4× indicates that demand for blockspace is being effectively absorbed by the rollup tier rather than displaced from the base layer.
07
Glossary & Methodology
Definitions and methodology references for figures presented in this report.
REVISION
G/03
Definitions
- Rollup
- An L2 protocol that posts compressed transaction data to a base layer (L1) for data availability and settlement.
- Optimistic Rollup
- A rollup that assumes validity by default and admits fault proofs during a challenge window.
- Validity (zk) Rollup
- A rollup that posts a succinct cryptographic proof attesting to the correctness of every state transition.
- Blob
- A 128 KiB data unit introduced by EIP-4844 used for cheap, ephemeral data availability.
Methodology
All figures are calculated as 7-day medians ending at D-0 unless otherwise stated. TPS is measured at the rollup’s execution layer rather than at the L1 inclusion event. Fees are computed as user-paid totals net of any sequencer rebates.
- ›Sample population: 47 rollups with continuous uptime exceeding 90 days.
- ›Currency: USD, converted at daily VWAP at 00:00 UTC.
- ›Source data: aggregated from public archive nodes and DA layer indexers.