Starknet Roadmap

Phase 1
Phase 2
Phase 3
Phase 4
Phase 5
Protocol features
  • Settlement on Ethereum
  • Native account abstraction
  • Cairo and Sierra
  • Parallel execution
  • Block packing 
  • Blobs for data availability
  • Fee payment in STRK 
  • 2-second tx confirmation
  • Sub-cent fees
  • <1 second pre-confirmations
  • Fee market
  • S-two prover is integrated into SHARP
  • Cairo Native
  • Bitcoin bridge – federated model, BitVM
  • 2-second block time
  • Full settlement on Bitcoin +  Ethereum
Performance
  • 5+ sustained TPS
  • 10 second + tx confirmation
  • 100+ sustained TPS
  • 400+ sustained TPS
  • 1k max TPS
  • 1k+ sustained TPS
  • 5k+ max TPS
  • 10k+ sustained TPS
Decentralization
  • Proof mechanism from day 1
  • Staking v1 – permissionless protocol and stake delegation
  • Open-source stack
  • First governance vote on Starknet mainnet
  • SN Stack
  • Stage 1 Rollup on Ethereum – L2Beat
  • Staking v2 – block attestation
  • Distributed sequencer architecture
  • Mempool
  • Staking v3 – decentralized block validation
  • Staking v4 – fully decentralized consensus and network operation
  • Decentralized proving
Product offerings
  • Starknet, the permissionless L2 validity rollup, is born on Ethereum
  • Stake STRK, delegate stake
  • Appchains
  • Bitcoin integrations on Starknet projects
  • Bitcoin wallets integration – Xverse
  • Bitcoin DeFi (BTCFi) on Starknet
  • BTC Staking
  • Asset Runes bridging
  • S-two verifier on Starknet
  • Client-side proving 
  • Trust-minimized BTC DeFi – making use of trust-minimized Bitcoin bridge
  • Fully trustless, native bridge between Starknet and Bitcoin
Phase 1
Protocol features
  • Settlement on Ethereum
  • Native account abstraction
  • Cairo and Sierra
Performance
  • 5+ sustained TPS
  • 10 second + tx confirmation
Decentralization
  • Proof mechanism from day 1
Product offerings
  • Starknet, the permissionless L2 validity rollup, is born on Ethereum
Phase 2
Protocol features
  • Parallel execution
  • Block packing 
  • Blobs for data availability
  • Fee payment in STRK 
  • 2-second tx confirmation
  • Sub-cent fees
Performance
  • 100+ sustained TPS
Decentralization
  • Staking v1 – permissionless protocol and stake delegation
  • Open-source stack
  • First governance vote on Starknet mainnet
  • SN Stack
Product offerings
  • Stake STRK, delegate stake
  • Appchains
Phase 3
Protocol features
  • <1 second pre-confirmations
  • Fee market
  • S-two prover is integrated into SHARP
  • Cairo Native
Performance
  • 400+ sustained TPS
  • 1k max TPS
Decentralization
  • Stage 1 Rollup on Ethereum – L2Beat
  • Staking v2 – block attestation
  • Distributed sequencer architecture
  • Mempool
Product offerings
  • Bitcoin integrations on Starknet projects
  • Bitcoin wallets integration – Xverse
  • Bitcoin DeFi (BTCFi) on Starknet
  • BTC Staking
  • Asset Runes bridging
Phase 4
Protocol features
  • Bitcoin bridge – federated model, BitVM
  • 2-second block time
Performance
  • 1k+ sustained TPS
  • 5k+ max TPS
Decentralization
  • Staking v3 – decentralized block validation
  • Staking v4 – fully decentralized consensus and network operation
Product offerings
  • S-two verifier on Starknet
  • Client-side proving 
  • Trust-minimized BTC DeFi – making use of trust-minimized Bitcoin bridge
Phase 5
Protocol features
  • Full settlement on Bitcoin +  Ethereum
Performance
  • 10k+ sustained TPS
Decentralization
  • Decentralized proving
Product offerings
  • Fully trustless, native bridge between Starknet and Bitcoin

Phase 1: The Launch

Done

Phase 1 began with the launch of Starknet Mainnet, a validity rollup that uses STARK proofs to ensure the integrity of every state update. From the outset, Starknet has required a valid proof for every change to the network’s state, making it impossible for invalid transactions to slip through.


Starknet transactions settle on Ethereum by verifying STARK proofs that attest to each state transition. This ensures the network inherits Ethereum’s security guarantees and decentralization. Because proofs must be verified on Ethereum, no entity—not even StarkWare or the Starknet Foundation—can manipulate user data or assets without detection. Starknet also launched with
native account abstraction, making smart-contract accounts the default and enabling developers to deliver Web2-style UX, such as passkey logins and biometric authentication.

Phase 2: Performance Meets Engagement

Done

Phase 2 focused on enhancing Starknet’s performance. The v0.13.2 ‘Bolt’ upgrade introduced parallel execution, enabling the simultaneous processing of independent transactions, thus increasing throughput. Block packing was also implemented, decoupling L2 block production from L1 state updates, leading to faster confirmations and lower L1 costs. Additionally, Starknet adopted Ethereum’s EIP-4844 upgrade, utilizing blobs for more efficient data availability. These enhancements enabled the network to reach several key performance milestones:

  • 127 sustained tps
  • 2-second transaction confirmations
  • sub-cent fees

Phase 2 also marked meaningful progress toward Starknet’s decentralization goals. While the network has been secured by a fully functioning validity proof mechanism since inception, the long-term vision is to evolve into a fully decentralized Proof-of-Stake (PoS) network. Two important pillars advanced during this phase were:

SN Stack

Key components of the Starknet infrastructure were made publicly available for projects looking to build their very own custom chains.


While this doesn’t directly decentralize public Starknet, it empowers others to build their own rollups or appchains using the same ZK tech stack that powers public Starknet. This broadens participation in the ecosystem and lays foundational groundwork for future decentralization.

Staking v1

STRK staking was introduced with a permissionless staking protocol and stake delegation, in which validators are required to run full nodes, in preparation for future validator responsibilities in the next staking stages. This gradual approach to implementing a staking mechanism allows for careful integration and gives network participants time to adapt. The vote to usher in the minting curve for staking rewards was the first community vote to be conducted directly on Starknet mainnet. 

Phase 3: On the Bitcoin-Aligned Path to Decentralization

In progress

This is Starknet’s current phase. It includes Starknet’s plan to become a unifying Layer 2 for both Ethereum and Bitcoin, and major advancements in decentralization.


Since announcing the campaign to become the execution layer for Bitcoin
in early 2025, with the intention of scaling Bitcoin from 13 tps to thousands and enabling complex functionality such as DeFi dApps, a host of Bitcoin-related product offerings have already been implemented on Starknet, including:

  • Lightning Network payments with the Braavos wallet
  • Integration of the Xverse Bitcoin wallet on Starknet
  • Atomiq provides cross-chain conversion solution for direct swap between native BTC and wBTC
  • A federated Runes bridge with Asset Runes

    As for decentralization, this phase both introduces more responsibilities to validators as part of the next stage of STRK staking, and also lays the groundwork for the full decentralization of Starknet, with a protocol upgrade implementing a distributed sequencer layer.

    Staking v2

    During the second phase of staking, validators gain more network-maintenance responsibility and begin attesting blocks, enabling public measurement of their performance and reliability. Staking v2 also strengthens the incentive structure for validators by allowing them to increase their commission from delegators. This protocol upgrade is a foundational upgrade in Starknet’s journey toward a more decentralized and economically sustainable protocol.

    Distributed sequencer architecture

    Starknet v0.14.0, the ‘Grinta’ upgrade,  marks a significant milestone on the road to full decentralization. With this upgrade, the network lays the foundation for a distributed sequencer layer: three independent sequencers reaching consensus and building blocks, each with its own mempool, and a native fee market to support future operator incentives. This version also introduces improved latency with sub-second pre-confirmations and a standardized paymaster interface for ease of integration.

    Phase 3 will also include major improvements to the network’s performance with the implementation of the open-source S-two prover into SHARP, the shared proving service, and the implantation of Cairo Native, which will allow for native smart contract execution. Both will bring about higher throughput and lower latency, as well as lower fees.

    Phase 4: Bitcoin Headway

    Comming soon

    While advocating for the implementation of Bitcoin’s OP_CAT upgrade, which will enable Starknet’s trustless settlement on Bitcoin, Starknet is also actively researching and implementing other Starknet-to-Bitcoin bridges, like additional federated models, BitVMs and the ColliderVM, which will open the door to more Bitcoin DeFi on Starknet.

     

    Other than advances on the Bitcoin front, Phase 4 will also align all the elements for full decentralization of the network with Starknet v0.15.0 and Staking v3 and v4:

    Starknet v0.15.0

    This version continues the network’s journey toward full decentralization by introducing decentralized block validation—ensuring that block correctness can be independently verified by anyone, not just trusted parties.

    Staking v3

    During this stage, validators will validate and vote on blocks sequenced by the sequencers. At this point, participation in consensus and block validation will be permissionless. A block will be finalized only if more than ⅔ of the stake voted on it.

    Staking v4

    Staking v4 will allow validators to assume full responsibility for maintaining and operating the network. However, the proving process will remain centralized under StarkWare.

    Lastly, a S-two verifier on Starknet is expected to make it the default network for verifying client-side proofs. S-two, the fastest prover in the world and the most suited for client-side proving, was introduced during Phase 3.

    Phase 5: Bitcoin and Decentralization Culmination

    Stay tuned

    The final phase of the Starknet roadmap will see Starknet’s staking protocol reach full decentralization with the decentralization of the proving mechanism. In addition, Starknet will operate as the unifying Layer 2, settling on both Bitcoin and Ethereum. Starknet will massively scale Bitcoin and natively open its community to complex use cases. As the execution layer of both Bitcoin and Ethereum, Starknet will enable the trustless flow of assets and a shared prosperity between the two largest blockchain communities.

     

    As for performance, we can expect to reach 10,000+ TPS during this final phase. And while Phase 5 is the end of Starknet’s current roadmap, it’s only the beginning of what’s to come.

     

    Stay tuned.

     

    Last updated: Aug 10, 2025