
Loading...

Loading...
Developer platform
One console for the whole SDK: run encrypted inference, chat, agents, batch jobs, worker operations, the bridge, and the DAO - each as a live, runnable panel, with the exact code to copy. Pick a capability on the left, or start below.
npm install lightnode-sdk viemnpm create lightnode-app my-appEncrypted end to end. ECDH + AES-GCM with the worker. The SDK never sees plaintext after it leaves the caller.
Verifiable on-chain. Every call leaves createSession + submitJob + jobCompleted receipts anyone can check.
Non-custodial. Your key signs locally. No accounts, no hosted middleman, single peer dep (viem).
Five lines to your first inference
import { runInferenceWithKey } from "lightnode-sdk";
const { answer, txs } = await runInferenceWithKey({
network: "testnet", // or "mainnet"
privateKey: process.env.PRIVATE_KEY, // 0x... funded key
prompt: "One fun fact about the ocean.",
});
console.log(answer); // decrypted answer
console.log(txs.createSession); // verifiable on-chain receiptsOne encrypted prompt to a verifiable answer
Pre-spend cost, worker depth, reliability, refund window
Multi-turn conversation, one job per turn
ReAct tool-calling loop on any model
Many prompts in parallel, stable order
Status, settle, stuck-job recovery, exit
Project worker earnings + ROI from live economics
Quote + move LCAI to/from Ethereum
Read + vote on LCAI Governor proposals
Live workers, models, jobs, analytics
Protocol fee revenue, run-rate, FeePool flow
Supply vs demand + saturation per model