BlockRun
Integrations / Vercel AI SDK

Use BlockRun from the Vercel AI SDK with the TypeScript client

Call every BlockRun model from a Vercel AI SDK app with the @blockrun/llm TypeScript client, or through an OpenAI-compatible provider pointed at a local x402 sidecar.

How it works

The @blockrun/llm TypeScript client creates or adopts a wallet, signs x402 payments and exposes chat, image, video and search calls with types. Inside a Vercel AI SDK app it is the shortest path: call the client from a route handler or a tool and stream the result back through the SDK's own primitives.

For code that must stay on the SDK's provider abstraction, the OpenAI-compatible provider pointed at a local blockrun-litellm sidecar gives the same catalog without a provider key; the sidecar signs each request from the wallet on the server.

Behind it is the BlockRun catalog: 78 chat models plus image, video, music and speech generation, search, market data and multi-chain RPC, priced per call. 6 open-weight models are free and need no wallet.

Setup

1. Install the client

npm install @blockrun/llm

2. Call a model

import { LLMClient } from "@blockrun/llm";
const client = new LLMClient();
const reply = await client.chat("openai/gpt-4o", "Hello!");

Questions

Can this run in the browser?
Keep the wallet server-side. Call the client from a route handler or server action and stream to the client.
Which chains?
Base by default; the client also supports paying on Solana.