Ecosystem
Multichain and interoperability
How AI jobs and applications can move across networks while preserving security and status transparency.
Multichain by design
BitcoinAI is designed to coordinate activity across multiple ecosystems.
- Bitcoin-connected applicationsResearch target
- Cosmos / IBCPlanned
- EthereumPlanned
- BasePlanned
- ArbitrumPlanned
- OptimismPlanned
- BNB ChainPlanned
- PolygonPlanned
- SolanaResearch target
- AppchainsPlanned
- Future chainsResearch target
No external-chain integration is live. Each route is listed with its own status.
- L3AI protocolJob, receipt, verification and settlement semantics
- L2Message formatTransport-agnostic cross-chain AI request
- L1TransportIBC, bridge adapters, messaging networks — per route
The AI transaction model should remain independent from the transport layer. IBC, bridge adapters, messaging networks, and future interoperability systems carry messages; they do not define the BitcoinAI AI protocol itself.
AI that crosses chains
An AI request can begin on one chain and return somewhere else.
Cross-chain AI jobs are asynchronous state machines with explicit request, acknowledgment, timeout, verification, and settlement states.
- 01Source Chain
- 02Cross-Chain Request
- 03BitcoinAI
- 04BatteryAGI
- 05Proof-of-AI
- 06BitcoinAI Verification
- 07Settlement / Callback
- 08Destination Chain
Explicit job states
- 01
REQUESTEDRequest emitted on the source chain - 02
RECEIVEDBitcoinAI acknowledges the packet - 03
ESCROWEDPayment locked against the job - 04
EXECUTINGBatteryAGI routes and runs the job - 05
VERIFIEDProof-of-AI receipt checked on BitcoinAI - 06
CALLBACK_PENDINGResult relayed to the destination - 07
CALLBACK_COMPLETEDestination acknowledged; settled FAILED / TIMEOUTFailure state: Deadline passed or proof invalid → refund path
Cross-chain jobs are asynchronous, not pretend atomic transactions across unrelated consensus systems.
Every route has a security model
Cross-chain connectivity must be evaluated route by route.
| Risk | Control |
|---|---|
| Bridge compromise | Control: Route-specific policy, value limits, circuit breakers |
| Message replay | Control: Sequence numbers and nonces bound to chain IDs |
| Stale state | Control: Timeouts and acknowledgments |
| Destination-chain reorg | Control: Finality thresholds per destination |
| Validator mismatch | Control: Light clients where available |
| Oracle failure | Control: Multiple sources, bounds and fallbacks |
| Callback failure | Control: Retry windows and refund paths |
Controls toolkit
- Sequence numbers
- Nonces
- Timeouts
- Acknowledgments
- Finality thresholds
- Light clients
- Route-specific policy
- Circuit breakers
What is ready vs. what is planned
Application capabilities should be transparent before mainnet.
- CanonicalPart of the current canonical protocol design. Not yet live; may change before mainnet through the published specification process.
- PlannedPart of the current design and roadmap. Not yet launched.
- Research targetAn engineering or research goal. Not a demonstrated production benchmark; subject to independent validation.
- EngineeringBeing specified and engineered. Exact parameters and rules are not final.
- Testnet TBDNot yet set. The value will be defined in the testnet specification and published before it applies.
| Capability | Scope | Status | |
|---|---|---|---|
| Sovereign L1 | Canonical | Cosmos SDK / CometBFT sovereign Layer 1 | Canonical |
| Smart contracts | Planned | Application-layer programmable state | Planned |
| EVM compatibility | Planned | EVM-compatible smart-contract execution | Planned |
| CosmWasm / WASM | Planned | WebAssembly smart contracts (CosmWasm) | Planned |
| IBC | Planned | IBC channels to Cosmos-ecosystem chains | Planned |
| EVM adapters | Planned | Adapters for Ethereum and EVM L2 ecosystems | Planned |
| Solana adapters | Research target | Adapters for Solana-connected applications | Research target |
| Appchain SDK | Planned | Tooling to connect appchains to BitcoinAI | Planned |
| Cross-chain AI request | Engineering | Asynchronous request / acknowledgement / timeout / callback | Engineering |
| BTC payment adapter | Research target | Escrow adapter, channel, bridge or L2 — not selected | Research target |
| Tokenized asset toolkit | Planned | Reference contracts and workflows for tokenized assets | Planned |
| NFT toolkit | Planned | Reference NFT / digital-asset contracts | Planned |
| AI marketplace | Planned | Composable AI services | Planned |
| Enterprise private-AI integration | Planned | Private execution with public commitments | Planned |
| Financial-market modules | Research target | Settlement, collateral and audit-trail primitives for licensed operators | Research target |
| Agent permission framework | Planned | Application-layer policy engine | Planned |
| Appchain templates | Planned | Starter appchain configurations | Planned |
| Developer SDK | Planned | @bitcoinai/* TypeScript packages — not published | Planned |
| Explorer | Planned | Block and AI-job explorer | Planned |
| Faucet | Planned | Test-token faucet | Planned |
| Public testnet | Testnet TBD | Launch date and parameters not yet set | Testnet TBD |
Nothing on this page is live on mainnet. Statuses update here as each capability ships.