IntegrationsVercel AI SDK

Vercel AI SDK

Use Tensoras.ai with the Vercel AI SDK to build AI-powered applications in Next.js, Nuxt, SvelteKit, and other frameworks.

Installation

npm install ai @tensoras/ai-provider

Authentication

Set your API key as an environment variable:

export TENSORAS_API_KEY="tns_your_key_here"

Or in your .env.local file for Next.js:

TENSORAS_API_KEY=tns_your_key_here

Provider Setup

import { tensoras } from "@tensoras/ai-provider";
 
const model = tensoras("llama-3.3-70b");

The provider reads TENSORAS_API_KEY from the environment. To pass it explicitly:

import { createTensoras } from "@tensoras/ai-provider";
 
const tensoras = createTensoras({
  apiKey: "tns_your_key_here",
});
 
const model = tensoras("llama-3.3-70b");

Generate Text

import { generateText } from "ai";
import { tensoras } from "@tensoras/ai-provider";
 
const { text } = await generateText({
  model: tensoras("llama-3.3-70b"),
  prompt: "Explain retrieval-augmented generation in one sentence.",
});
 
console.log(text);

Stream Text

import { streamText } from "ai";
import { tensoras } from "@tensoras/ai-provider";
 
const result = streamText({
  model: tensoras("llama-3.3-70b"),
  prompt: "Write a haiku about open-source AI.",
});
 
for await (const chunk of result.textStream) {
  process.stdout.write(chunk);
}

Generate Object (Structured Output)

import { generateObject } from "ai";
import { tensoras } from "@tensoras/ai-provider";
import { z } from "zod";
 
const { object } = await generateObject({
  model: tensoras("llama-3.3-70b"),
  schema: z.object({
    recipe: z.object({
      name: z.string(),
      ingredients: z.array(z.string()),
      steps: z.array(z.string()),
    }),
  }),
  prompt: "Generate a recipe for chocolate chip cookies.",
});
 
console.log(object.recipe.name);
console.log(object.recipe.ingredients);

Tool Calling

import { generateText, tool } from "ai";
import { tensoras } from "@tensoras/ai-provider";
import { z } from "zod";
 
const result = await generateText({
  model: tensoras("llama-3.3-70b"),
  tools: {
    weather: tool({
      description: "Get the current weather for a city",
      parameters: z.object({
        city: z.string().describe("The city name"),
      }),
      execute: async ({ city }) => {
        return { city, temperature: 72, condition: "sunny" };
      },
    }),
  },
  prompt: "What is the weather in San Francisco?",
});
 
console.log(result.text);

Next.js: Chat Route

Create an API route that streams responses:

// app/api/chat/route.ts
import { streamText } from "ai";
import { tensoras } from "@tensoras/ai-provider";
 
export async function POST(req: Request) {
  const { messages } = await req.json();
 
  const result = streamText({
    model: tensoras("llama-3.3-70b"),
    system: "You are a helpful assistant.",
    messages,
  });
 
  return result.toDataStreamResponse();
}

Next.js: Chat UI with useChat

Build a chat interface using the useChat hook:

// app/page.tsx
"use client";
 
import { useChat } from "ai/react";
 
export default function Chat() {
  const { messages, input, handleInputChange, handleSubmit, isLoading } =
    useChat();
 
  return (
    <div>
      <div>
        {messages.map((m) => (
          <div key={m.id}>
            <strong>{m.role === "user" ? "You" : "AI"}:</strong>
            <p>{m.content}</p>
          </div>
        ))}
      </div>
 
      <form onSubmit={handleSubmit}>
        <input
          value={input}
          onChange={handleInputChange}
          placeholder="Type a message..."
          disabled={isLoading}
        />
        <button type="submit" disabled={isLoading}>
          Send
        </button>
      </form>
    </div>
  );
}

Next.js: RAG with Knowledge Bases

Use Tensoras Knowledge Bases in your API route for RAG:

// app/api/chat/route.ts
import { streamText } from "ai";
import { tensoras } from "@tensoras/ai-provider";
import Tensoras from "tensoras";
 
const tns = new Tensoras();
 
export async function POST(req: Request) {
  const { messages } = await req.json();
  const lastMessage = messages[messages.length - 1].content;
 
  // Retrieve relevant context from Knowledge Base
  const retrieval = await tns.chat.completions.create({
    model: "llama-3.3-70b",
    messages,
    knowledge_bases: ["kb_a1b2c3d4"],
    stream: false,
  });
 
  const context = retrieval.citations
    ?.map((c) => c.text)
    .join("\n\n");
 
  const result = streamText({
    model: tensoras("llama-3.3-70b"),
    system: `You are a helpful assistant. Use the following context to answer questions:\n\n${context}`,
    messages,
  });
 
  return result.toDataStreamResponse();
}

Embedding Model

Use Tensoras embeddings with the Vercel AI SDK:

import { embed } from "ai";
import { tensoras } from "@tensoras/ai-provider";
 
const { embedding } = await embed({
  model: tensoras.textEmbeddingModel("gte-large-en-v1.5"),
  value: "What is deep learning?",
});
 
console.log(`Dimensions: ${embedding.length}`);

Next Steps

CrewAIHaystack