🔥BNB Chain's new Layer 1: Expanding the ecosystem or competing with itself?

🔴BNB Chain has just announced its H2/2026 roadmap, focusing on a brand new layer-1 for high-frequency AI Agent transactions. Testnet end of 2026, mainnet early 2027.
🔴Target specifications:
- Over 100,000 TPS via parallel execution, with a long-term target of 1 million TPS
- Pre-confirmation under 50ms, finality under 1 second
- TxStream: remove public mempool, block MEV and front-running from the root
- PriorityLane: dedicated bandwidth for oracles, bridges, liquidations
🔴Will the new chain drain liquidity from BSC or expand the pie?
It is highly likely that the L1 will be a sidechain of BNB, with 3 signals:
- CZ said BSC remains the settlement hub, the new chain connects via native bridge
- BNB is the unified asset across all chains: gas fees, staking all flow to one token
- Clear layering: BSC handles general DeFi, the new chain handles HFT and AI Agent - use cases BSC was not originally designed for
That means BNB Chain is following a specialized multi-chain playbook: Each chain has its own purpose, with a common token that accumulates value. The complete opposite of the Solana scenario, where everything is concentrated on a single chain.
🔴Nevertheless, there will be a period of self-competition. DEX perps and trading bots - the biggest fee contributors on BSC - are exactly the cohort that will migrate to the new chain first. In the short term, BSC will lose volume; long term, value still accrues to BNB if the bridge and experience are smooth enough.
There are clear lessons within the system itself: opBNB and Greenfield were also promoted as expansion pieces but never gained their own traction. Let's see what CZ builds this time.