Key Notes
- The Ethereum Layer-2 solution upgraded capacity after successfully transitioning from op-geth to op-reth client infrastructure.
- Network execution speed remains the primary scaling constraint requiring resolution before achieving higher throughput targets.
- Plans include reaching 400-500 Mgas/s by early 2026 through TrieDB implementation delivering significantly faster state retrieval.
Base founder Jesse Pollak announced on Nov. 6 that the network increased its gas limit from 100 to 125 million gas per second. The capacity increase moves the Ethereum Layer-2 closer to its stated year-end target of 150 Mgas/s.
NEWS: just scaled @base from 100 to 125 Mgas/s pic.twitter.com/PR0RcgkNhV
— jesse.base.eth (@jessepollak) November 6, 2025
Base published an engineering blog post on Oct. 28 outlining its scaling roadmap. The post, authored by engineer Anika Raghuvanshi, committed to doubling the network’s gas limit from 75 to 150 Mgas/s by the end of 2025.
Migration to Reth Client Enables Capacity Jump
The network completed its migration from op-geth to the op-reth client software over recent months. Base migrated its sequencer nodes to Reth most recently, according to the engineering post.
The team measured the Reth client as significantly more performant than the previous op-geth client. Base now recommends external validators use Reth as the default client moving forward.
Execution Speed Identified as Primary Bottleneck
Base identified client execution speed as its biggest current scaling bottleneck. The Base network explores plans for a potential native token while addressing infrastructure constraints. The team previously resolved limitations around L1 data availability and fault-proof performance.
The engineering team built a benchmarking tool to simulate block building times at specified gas limits with various traffic patterns. The tool illuminated specific performance constraints that needed addressing before further scaling.
Future Scaling Plans Target 400-500 Mgas/s
Base set a target of 400-500 Mgas/s by early 2026. The target depends on completing the TrieDB database project and implementing new resource metering tools.
TrieDB restructures the database format to make fetching state faster. The team stated it is close to having a final version of the project, which it expects to deliver 8-10x faster storage reads.
The scaling efforts aim to maintain transaction fees below one cent. The network saw fees reach five cents during periods of high activity in June 2025. Recent deployments including XSwap’s token creation platform and Stripe’s USDC subscription payments add to application activity on the network.
next
