Overview

Linera is a decentralized protocol optimized for real-time, agentic Web3 applications that require guaranteed performance for an unlimited number of active users.

The core idea of the Linera protocol is to run many chains of blocks, called microchains, in parallel in one set of validators.

How do microchains work?

Linera users propose blocks directly to the chains that they own. Chains may also be shared with other users. Linera validators ensure that all blocks are validated and finalized in the same way across all the chains.

    flowchart LR
        user(["User wallet"])
        provider(["Provider"])
        validators(["Validators"])
        subgraph "Validators"
          chain1["Personal chain"]
          chain2["Temporary chain"]
          chain3["Public chain"]
          chain4["App chain"]
        end
        user -- owns --> chain1
        user -- shares --> chain2
        validators -- own --> chain3
        provider -- owns/shares --> chain4

    %% Styling
    style Validators fill:#1A4456,stroke:#70D4D3,stroke-width:2px,stroke-dasharray:3 3,rx:10,ry:10
    style user fill:#8B7355,stroke:#EDE4D2,stroke-width:2px
    style provider fill:#A0736B,stroke:#D2E8C8,stroke-width:2px
    style chain1 fill:#3A6B7A,stroke:#A0E3E2,stroke-width:2px
    style chain2 fill:#5A8269,stroke:#D2E8C8,stroke-width:2px
    style chain3 fill:#4A5A60,stroke:#F3EEE2,stroke-width:2px
    style chain4 fill:#4A7B75,stroke:#70D4D3,stroke-width:2px

While validation rules and security assumptions are the same for all chains, block production in each chain can be configured in a number of ways. In practice, most chains fall into the following categories:

• Personal chains (aka. user chains”) are those with a single owner, i.e. a single user proposing blocks.
• Temporary chains are shared between a few users.
• Public chains, usually dedicated to a particular task in the Linera infrastructure, are fully managed by Linera validators.
• Chains dedicated to a particular application, called app chains, may use either their own infrastructure for block production, a permissionless solution using proof-of-work, or a rotating set of trusted providers.

In order to validate all the chains reliably and at minimal cost, Linera validators are designed to be elastic, meaning that they can independently add or remove computational power (e.g. cloud workers) on demand whenever needed. In turn, this allows Linera applications to scale horizontally by distributing work to the microchains of their users.

What makes Linera real-time and agent-friendly?

Connected clients

To propose blocks and provide APIs to frontends, Linera users rely on a Linera client. Clients synchronize on-chain data in real-time, without trusting third parties, thanks to local VMs and Linera’s native support for notifications. Clients are sparse in the sense that they track only the chains relevant to a particular user.

flowchart LR
    subgraph user_device["User device"]
        ui["Web UI or AI agent"]
        ui <-- GraphQL --> user_local

        subgraph linera_client["Linera client"]
            user_local["user chain"]
            admin_local["admin chain"]
        end
    end

    user_local <-- sync blocks --> user_remote
    user_remote -- notify incoming messages --> user_local
    admin_local <-- sync blocks --> admin_remote

    subgraph validators_clients["Validators"]
        user_remote["user chain"]
        admin_remote["admin chain"]
    end

    %% Styling
    style user_device fill:#1A4456,stroke:#70D4D3,stroke-width:2px,stroke-dasharray:3 3,rx:10,ry:10
    style linera_client fill:#0e2630,stroke:#A0E3E2,stroke-width:1px,stroke-dasharray:2 2,rx:8,ry:8
    style validators_clients fill:#1A4456,stroke:#70D4D3,stroke-width:2px,stroke-dasharray:3 3,rx:10,ry:10

    style ui fill:#8B7355,stroke:#EDE4D2,stroke-width:2px
    style user_local fill:#3A6B7A,stroke:#A0E3E2,stroke-width:2px
    style admin_local fill:#3A6B7A,stroke:#A0E3E2,stroke-width:2px
    style user_remote fill:#4A5A60,stroke:#F3EEE2,stroke-width:2px
    style admin_remote fill:#4A5A60,stroke:#F3EEE2,stroke-width:2px

User interfaces interact with Linera applications by querying and sending high-level commands to local GraphQL services running securely inside the Linera client.

Similarly, discussions between AI agents and Linera applications stay local, hence private and free of charge. This also protects agents against compromised external RPC services.

Linera is the first Layer-1 to allow trustless real-time synchronization of user data on their devices, democratizing low-latency data access and bringing professional-grade security to frontends, customized oracle networks, and AI-trading agents.

Geographic sharding

In the future, Linera validators will be incentivized to operate machines and maintain a presence in a number of key regions. Most microchains will be pinned explicitly to a specific region, giving users of this region the lowest latency possible in their on-chain interactions. Linera validators will be incentivized to connect their regional data-centers using a low-latency network.

flowchart LR
    subgraph region_1["Region 1"]
        user1(["User 1"])
        chain1["chain 1"]
        chain2["chain 2"]
    end
    subgraph region_2["Region 2"]
        chain3["chain 3"]
        chain4["chain 4"]
    end
    user1 --> chain1
	  chain1 <--> chain2
	  chain2 <---> chain3
	  chain2 <---> chain3
	  chain3 <--> chain4

    %% Styling
    style region_1 fill:#1A4456,stroke:#70D4D3,stroke-width:2px,stroke-dasharray:3 3,rx:10,ry:10
    style region_2 fill:#1A4456,stroke:#70D4D3,stroke-width:2px,stroke-dasharray:3 3,rx:10,ry:10
    style user1 fill:#8B7355,stroke:#EDE4D2,stroke-width:2px
    style chain1 fill:#3A6B7A,stroke:#A0E3E2,stroke-width:2px
    style chain2 fill:#3A6B7A,stroke:#A0E3E2,stroke-width:2px
    style chain3 fill:#4A5A60,stroke:#F3EEE2,stroke-width:2px
    style chain4 fill:#4A5A60,stroke:#F3EEE2,stroke-width:2px

Importantly, geographic affinity in Linera is not conditioned to the time of the day allowing applications to deliver similar performance at night and during working hours. Yet, Linera validators have the flexibility to downsize and upsize their pool of machines at will to save costs.

How do Linera microchains compare to traditional multi-chain protocols?

Linera is the first blockchain designed to run a virtually unlimited number of chains in parallel, including one dedicated user chain per user wallet.

In traditional multi-chain protocols, each chain usually runs a full blockchain protocol in a separate set of validators. Creating a new chain or exchanging messages between chains is expensive. As a result, the total number of chains is generally limited.

In contrast, Linera is designed to run as many microchains as needed:

• Users only create blocks in their chain when needed;

• Creating a microchain does not require onboarding validators;

• All chains have the same level of security;

• Microchains communicate efficiently using the internal networks of validators;

• Validators are internally sharded (like a regular web service) and may adjust their capacity elastically by adding or removing internal workers.

• Users may run heavy transactions in their microchain without affecting other users.

Main protocol features

Infrastructure

• Finality time under 0.5 seconds for most blocks, including a certificate of execution.

• New microchains created in one transaction from an existing chain.

• No theoretical limit in the number of microchains, hence the number of transactions per second (TPS).

• Bridge-friendly block headers compatible with EVM signatures

On-chain applications

• Rich programming model allowing applications to distribute computation across chains using asynchronous messages, shared immutable data, and event streams.

• Full synchronous composability inside each microchain.

• Support for heavy (multi-second) transactions and direct oracle queries to external web services and data storage layers.

Web client and wallet infrastructure

• Real-time push-notifications from validators to web clients.

• Block synchronization and VM execution for selected microchains, allowing instant pre-confirmation of user transactions.

• Trustless reactive programming using familiar Web2 frameworks.

• On-chain applications programmed in Rust to run on Wasm, or Solidity on EVM(*).

Features marked with (*) are under active development on the main branch.