Phase 1: Foundational layer for USDT
Status: Live on mainnet.StableBFT: Live
A customized PoS consensus protocol built on CometBFT. Delivers deterministic finality and Byzantine fault tolerance up to one-third of validators. See Consensus for the current implementation.USDT as native gas: Live
USDT0 is the native asset for gas payment and value transfer, and simultaneously supports the ERC-20 surface (approve, transfer, transferFrom, permit). See USDT as gas.
Stable Pay & Stable Name: In progress
Stable Pay is a Web2.5 UX wallet experience designed to simplify onboarding for new users while remaining compatible with existing Web3 wallets. Stable Name is a user-friendly aliasing system that replaces raw EVM addresses with human-readable identifiers for sending and receiving tokens.Phase 2: Experience layer for USDT
Status: Mixed. State DB optimization is live; the remaining items are in development.State DB optimization: Live
Stable decouples state commitment from state storage. Validators commit the latest state in memory while historical state is deferred to disk.MemDB and VersionDB powered by mmap handle real-time commitment without blocking on disk I/O. See Storage (StableDB).
Optimistic parallel execution: Planned
Real-world telemetry shows 60–80% of transactions interact with disjoint state and can safely execute in parallel. Stable will execute transactions optimistically under the no-conflict assumption, with rollback and sequential re-execution on detected conflicts. This preserves correctness while improving throughput.USDT Transfer Aggregator: Planned
Aggregating mechanism that groups USDT0 transfers and processes them collectively, reducing per-transaction overhead and improving overall throughput. See USDT transfer aggregator.Guaranteed blockspace: Planned
Reserved block capacity for enterprise partners, enforced through validator-level reservations and dedicated RPC endpoints. Delivers predictable latency for mission-critical payment flows even under network congestion. See Guaranteed blockspace.Phase 3: Full-stack optimized layer for USDT
Status: Planned.StableBFT on Autobahn
DAG-based BFT consensus that integrates naturally with Stable’s CometBFT-based consensus layer. See StableBFT for the current protocol and Autobahn for the target architecture. Internal proof-of-concept has demonstrated over 200,000 TPS (consensus only) in controlled environments.StableVM++
High-performance execution engine that replaces the Go-based EVM with a C++ implementation. Projected to deliver up to a 6x improvement in EVM execution speed.High performance RPC
A full RPC stack covering node-level enhancements (real-time chain state processing), node-integrated indexing (low-latency application APIs), scalable pub/sub over WebSocket, and a hybrid load balancer that routes by operation type. See High performance RPC for the current split-path architecture.Next recommended
Architecture overview
Walk through the current state of the stack the roadmap evolves.
Tokenomics
Review the economic model that funds validator incentives across the roadmap.
Tech overview
Return to the architecture summary.