BidBlock V2 Packs 28% More Gas Per Block
The most significant early result from Pasteur is the amount of gas being processed inside BidBlock V2 blocks.
According to BNB Chain, BidBlock V2 is averaging 28% more gas per block than Bid V1 under live mainnet traffic.
At the upper end of the distribution, BidBlock V2 reached approximately 43.8 million gas at P99, compared with 33.8 million for Bid V1.
These figures are based on real mainnet activity rather than a controlled benchmark, meaning they reflect observed production performance and should not be interpreted as an isolated measurement of BEP-675 alone.
Even so, the results indicate that BSC's new block-building architecture is allowing builders to construct deeper blocks.
BEP-675 Becomes the Dominant Block-Building Path
Adoption of the new architecture has also been rapid.
During the three-day observation period ending September 1, BNB Chain reported that BidBlock V2 accounted for approximately 98% of blocks.
Bid V1 represented around 2%, while locally produced blocks accounted for roughly 0.1%.
BNB Chain also reported that all 33 builders actively producing blocks during the observation period had integrated BEP-675.
That means the new architecture moved from a newly activated block-building option to BSC's primary production path within days of the Pasteur upgrade.
How Pasteur Changes BSC Block Production
The main improvement introduced by BEP-675 is relatively straightforward: less duplicated work before a block is sealed.
Previously, builder-proposed blocks could involve three EVM execution passes.
A validator imported the parent block, a builder executed transactions while constructing its bid, and the validator then re-executed those transactions before sealing the winning block.
BEP-675 removes that final validator-side replay from the pre-seal process.
Builders can instead submit a fully executed BidBlock containing transactions and execution results. Validators can then assemble and seal the block without repeating every user transaction first.
The change reduces the builder-driven path from three EVM execution passes to two.
Validators still execute and verify the block after it is broadcast, checking elements including the state root and receipts root.
Bid V1 also remains available as a fallback if a builder encounters problems using BidBlock V2.
BSC Moves Toward Higher Gas Limits
The additional block-building headroom is intended to support higher gas limits.
BSC entered September with a 55 million gas limit, but BNB Chain has outlined a staged path toward significantly larger blocks.
The first target is 70 million gas, followed by potential increases toward 80 million and 90 million.
These should not be treated as automatic upgrades.
BNB Chain says each increase will depend on evidence from live mainnet performance.
Important metrics include:
Block finality
Missed blocks
Large-block behavior
Cross-region latency
Node import times
Builder market share
Fallback-system health
This allows BSC to increase capacity gradually while monitoring whether validators, builders and network infrastructure can reliably handle larger blocks.
Larger Blocks Require More Infrastructure
Increasing the gas limit affects more than block construction.
Larger blocks must reach validators quickly, synchronize between nodes and be executed efficiently after they arrive.
BNB Chain is therefore working on supporting infrastructure alongside BEP-675.
One area involves using RLP-encoded data over gRPC to reduce serialization and representation overhead when larger BidBlocks are delivered.
Another development, eth/70, is designed to improve synchronization by allowing receipts for a single block to be delivered across multiple bounded responses.
BNB Chain is also researching Block-Level Access Lists, which could expose state access patterns in advance and support state prefetching, parallel execution and faster block imports.
However, BNB Chain says Block-Level Access Lists are not currently required for the planned move toward 90 million gas.
Why Pasteur Matters for BNB Chain
Additional block capacity could become increasingly important as activity expands across BSC.
DeFi protocols, decentralized exchanges, stablecoin transfers, trading applications and other high-volume dApps all compete for block space.
Higher gas limits allow more computational work to fit inside each block.
Pasteur does not simply increase the gas limit directly. Instead, it changes the architecture underneath block production to create more headroom for larger blocks.
That distinction is important because BNB Chain is attempting to increase capacity while continuing to monitor finality, synchronization and node performance.
BNB Chain Expands Tokenized Stock Development
The scaling work comes as BNB Chain is also encouraging developers to build new types of financial applications on BSC.
The ongoing BNB Hack: Tokenized Stocks Edition, organized with Binance Web3 Wallet, runs from September 16 to October 11, 2026 and offers a $20,000 prize pool.
Developers are being challenged to build products and agents around tokenized equities on BSC.
Potential projects include trading interfaces, portfolio tools, autonomous agents, rebalancing systems, dashboards and applications comparing onchain tokenized-stock prices with traditional market reference prices.
Participants also receive access to tools covering market data, RWA data, swap routing, transaction simulation, wallet information and DeFi positions.
The program shows how BNB Chain's infrastructure improvements are occurring alongside efforts to expand BSC into areas beyond traditional DeFi.
What Comes Next for BSC
The next phase will focus on how BSC performs as its gas capacity increases.
BidBlock V2 has already become the dominant block-building path and is producing deeper blocks under live network conditions.
The challenge now moves toward the rest of BSC's infrastructure.
Builders must deliver larger blocks within strict time limits, nodes must synchronize increasingly large payloads, and validators must continue importing and verifying those blocks efficiently.
BNB Chain's approach is therefore incremental: increase capacity, measure real-world performance and use that data to determine whether the network is ready for the next step.
Conclusion
BNB Smart Chain has gained additional block-processing headroom following the Pasteur hard fork, with BidBlock V2 averaging approximately 28% more gas per block than Bid V1 under observed mainnet conditions.
BEP-675 has also seen rapid adoption, accounting for approximately 98% of blocks during BNB Chain's latest three-day observation period.
The network is now moving toward higher gas limits, beginning with 70 million and potentially progressing to 80 million and 90 million as infrastructure and mainnet performance allow.
Pasteur therefore represents more than a single capacity increase. By removing duplicated execution from the critical block-building path, BSC has changed its underlying architecture to create additional room for future scaling.
BNB Chain — Post-Pasteur Hard Fork Technical Report
BNB Chain — BEP-675: Builder-Proposed Block Specification
BNB Chain — Pasteur Upgrade Specification (BEP-673)
BNB Chain — Tokenized Stocks Hackathon
