> ## Documentation Index > Fetch the complete documentation index at: https://docs.lucid.foundation/llms.txt > Use this file to discover all available pages before exploring further. # Run inference (optionally streaming via SSE) > Execute inference through the LucidLayer execution gateway. Supports both streaming (SSE) and non-streaming responses. A cryptographic receipt is generated for each successful inference. ## OpenAPI ````yaml /openapi-spec.yaml post /v1/run/inference openapi: 3.0.3 info: title: LucidLayer API version: 1.0.0 description: > LucidLayer API — the verifiable AI execution layer. Blockchain-anchored infrastructure giving AI assets (models, agents, tools, compute, datasets) provable identity, cryptographic receipts, and reputation backed by real traffic data. **Open-source and self-hostable.** Run your own instance with `docker compose up` or `npm start` (port 3001). The SDK defaults to `http://localhost:3001` — no cloud account required. This OpenAPI spec is the source of truth for all backend routes and is used to generate SDK clients via [Speakeasy](https://speakeasy.com). ## Version Strategy | Prefix | Status | Description | |--------|--------|-------------| | `/v1/*` | **Stable** | Core API — passports, receipts, epochs, payouts, inference, agent deployment. Production-ready and covered by semver guarantees. | | `/v2/*` | **Stable** | Cross-chain and reputation APIs — EVM multi-chain, ERC-8004, pluggable reputation. Production-ready. | | `/api/agents/*` | **Deprecated** | Legacy agent MMR orchestration. Retained for SDK backward compatibility only. Migrate to `/v1/agents/*` (AgentDeploy, AgentMirror). Will be removed in a future major version. | | `/health/*` | **Stable** | System health probes — not versioned (infrastructure endpoints). | ### Migration Path (legacy `/api/agents/*` to `/v1/agents/*`) 1. Replace `init_agent` + `process_agent_epoch` with `lucid_deploy_agent` (one-click deployment). 2. Replace `generate_agent_proof` with `lucid_get_agent_proof` / `lucid_get_agent_run_proof` (AgentMirror). 3. Replace `get_agent_stats` / `get_agent_history` with `lucid_get_agent_receipts` + `lucid_get_agent_epoch`. 4. All legacy operationIds will be removed when the next major version ships. ## Route Groups - `/v1/*` — Passports, Match, Run, Receipts, Epochs, Payouts, Compute - `/v1/agents/*` — Agent Deployment, Wallet, Revenue, Mirror (receipts/proofs) - `/v1/a2a/*` — A2A Protocol (Agent-to-Agent) - `/v1/config/*` — Payment configuration and grants - `/v2/chains/*` — Cross-chain status and routing - `/v2/agents/*` — ERC-8004 agent registration and reputation - `/v2/reputation/*` — Pluggable reputation scoring, cross-chain aggregation - `/v2/payouts/*` — On-chain payout execution - `/health/*` — System health and dependency checks - `/api/agents/*` — Legacy agent MMR orchestration (deprecated) servers: - url: http://localhost:3001 description: Local / self-hosted (default for open-source users) x-speakeasy-server-id: local - url: https://{host}:{port} description: Self-hosted instance (custom host and port) x-speakeasy-server-id: self-hosted variables: host: default: localhost description: Hostname or IP of your Lucid instance port: default: '3001' description: Port your Lucid instance listens on - url: https://api.lucid.foundation description: Managed cloud (Raijin Labs hosted) x-speakeasy-server-id: cloud security: - BearerAuth: [] - {} tags: - name: Passports description: >- AI asset identity — create, list, update, delete model/compute/tool/dataset/agent passports - name: Match description: Policy-based compute matching and route planning - name: Run description: OpenAI-compatible inference (chat completions) and model listing - name: Receipts description: Cryptographic receipts — create, verify, and generate MMR inclusion proofs - name: Epochs description: Epoch lifecycle — creation, finalization, on-chain anchoring, and retry - name: Payouts description: Revenue split calculation and payout execution (basis-point math) - name: Compute description: Compute node heartbeat registration and health monitoring - name: Health description: System health — liveness, readiness, and per-dependency status checks - name: Agents description: >- DEPRECATED: Legacy agent MMR orchestration (use AgentDeploy, AgentMirror instead) - name: Shares description: Share token launch and revenue airdrop for fractional AI asset ownership - name: Escrow description: Cross-chain escrow lifecycle — create, release, dispute, and query - name: Disputes description: Escrow dispute resolution — open, submit evidence, resolve, appeal - name: Paymaster description: >- ERC-4337 gas sponsorship — sponsor transactions, estimate fees, check rates - name: Identity description: >- Cross-chain identity bridging — CAIP-10 linking, resolution, and chain discovery - name: TBA description: ERC-6551 Token Bound Accounts — create and query agent wallets - name: Modules description: >- ERC-7579 module management — install, uninstall, and configure policy/payout modules - name: zkML description: >- Zero-knowledge ML — prove inference, verify proofs, register models on-chain - name: Payments description: >- x402 payment configuration — pricing, facilitators, grants, and subscription - name: AgentDeploy description: >- Agent deployment lifecycle — deploy, status, logs, scale, terminate across 6 targets - name: AgentWallet description: Agent PDA wallet — balance, send, policy limits, and transaction history - name: AgentRevenue description: >- Agent revenue pools — earnings tracking, airdrop triggers, and pool queries - name: A2A description: Google A2A protocol — agent cards, task submission, and task lifecycle - name: Reputation description: >- Pluggable reputation scoring — algorithms, compute scores, and composite aggregation - name: CrossChain description: >- EVM multi-chain operations — chain status, ERC-8004 registration, cross-chain reputation and payouts - name: AgentMirror description: >- Read-only agent proof endpoints — receipts, epochs, and MMR proofs (admin auth required) - name: Memory description: >- Agent memory — episodic, semantic, procedural, entity, trust-weighted, temporal types with hash-chain provenance - name: Anchoring description: >- DePIN anchoring control plane — query, verify, and trace artifact anchor records - name: AgentLaunch description: >- Agent activation — launch via image or base runtime, blue-green rollout, promotion, and rollback - name: Webhooks description: >- Deployment webhook receivers — normalizes provider-specific callbacks into deployment events paths: /v1/run/inference: post: tags: - Run summary: Run inference (optionally streaming via SSE) description: > Execute inference through the LucidLayer execution gateway. Supports both streaming (SSE) and non-streaming responses. A cryptographic receipt is generated for each successful inference. operationId: lucid_run_inference requestBody: required: true content: application/json: schema: $ref: '#/components/schemas/InferenceRequest' responses: '200': description: Non-streaming inference result content: application/json: schema: $ref: '#/components/schemas/InferenceResult' '422': $ref: '#/components/responses/UnprocessableEntity' '500': $ref: '#/components/responses/InternalServerError' '503': $ref: '#/components/responses/ServiceUnavailable' x-codeSamples: - lang: typescript label: Typescript (SDK) source: |- import { RaijinLabsLucidAi } from "raijin-labs-lucid-ai"; const raijinLabsLucidAi = new RaijinLabsLucidAi(); async function run() { const result = await raijinLabsLucidAi.run.inference({}); console.log(result); } run(); components: schemas: InferenceRequest: type: object properties: model_passport_id: type: string model: type: string prompt: type: string messages: type: array items: $ref: '#/components/schemas/ChatMessage' max_tokens: type: integer temperature: type: number top_p: type: number top_k: type: integer stop: type: array items: type: string stream: type: boolean policy: $ref: '#/components/schemas/Policy' compute_passport_id: type: string trace_id: type: string request_id: type: string InferenceResult: type: object required: - success - run_id - tokens_in - tokens_out - ttft_ms - total_latency_ms - model_passport_id - compute_passport_id - runtime properties: success: type: boolean run_id: type: string request_id: type: string trace_id: type: string text: type: string finish_reason: type: string tokens_in: type: integer tokens_out: type: integer ttft_ms: type: integer total_latency_ms: type: integer model_passport_id: type: string compute_passport_id: type: string runtime: type: string policy_hash: type: string receipt_id: type: string used_fallback: type: boolean fallback_reason: type: string error: type: string error_code: type: string ChatMessage: type: object required: - role - content properties: role: type: string enum: - system - user - assistant - function content: type: string name: type: string function_call: type: object properties: name: type: string arguments: type: string Policy: type: object required: - policy_version additionalProperties: false properties: policy_version: type: string description: Policy schema version, e.g. '1.0' allow_regions: type: array items: type: string residency_required: type: boolean attestation: type: object additionalProperties: false properties: attestation_required: type: boolean require_cc_on: type: boolean fallback_allowed: type: boolean latency: type: object additionalProperties: false properties: p95_ms_budget: type: integer hard_timeout_ms: type: integer cost: type: object additionalProperties: false properties: max_price_per_1k_tokens_usd: type: number spot_only: type: boolean privacy: type: object additionalProperties: false properties: store_inputs: type: boolean redact_pii: type: boolean ErrorResponse: type: object properties: success: type: boolean error: type: string message: type: string error_code: type: string details: {} responses: UnprocessableEntity: description: Unprocessable Entity content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' InternalServerError: description: Internal Server Error content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' ServiceUnavailable: description: Service Unavailable content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' securitySchemes: BearerAuth: type: http scheme: bearer description: Lucid API key (lk_live_... or lk_test_...) ````