Vercel AI SDK adapter

Package: @turbomem/vercel-ai

Exposes turbomem as two tools the model can call: remember facts and recall memories by semantic search.

Install

npm install @turbomem/vercel-ai turbomem ai

Usage

import { generateText } from "ai";
import { openai } from "@ai-sdk/openai";
import { TurboMemory } from "turbomem";
import { createMemoryTools } from "@turbomem/vercel-ai";

const memory = new TurboMemory({
  embeddings: "openai",
  storage: "pglite",
  extraction: { provider: "openai", model: "gpt-4o-mini" },
  openai: { apiKey: process.env.OPENAI_API_KEY },
});

await memory.init();

const result = await generateText({
  model: openai("gpt-4o-mini"),
  tools: createMemoryTools(memory, { userId: "user_123" }),
  prompt: "Remember that I prefer dark mode, then tell me my preferences.",
  maxSteps: 5,
});

createMemoryTools accepts a scope object (userId, agentId, sessionId) that is applied to every tool invocation.

See the vercel-ai-chatbot example for a full Next.js App Router integration.