Vercel x DeepInfra Chatbot

Vercel x DeepInfra Chatbot is an open-source AI chatbot application template that pairs Next.js with the AI SDK by Vercel and DeepInfra as the model provider, giving developers a working streaming chat app to clone, deploy, or extend.

What is Vercel x DeepInfra Chatbot?

Vercel x DeepInfra Chatbot is an open-source AI chatbot template that runs on Next.js using the App Router, streams model responses through the AI SDK by Vercel, and calls DeepInfra for inference. It takes user chat prompts as input and returns streaming text responses, optionally invoking tools and reasoning models. The repository is maintained by the Vercel team and community contributors, and a hosted demo runs at ai-sdk-starter-deepinfra.vercel.app.

Key Features

• Streaming responses via the AI SDK — text is streamed to the browser, and multiple AI providers can be swapped in interchangeably with a few lines of code.
• Built-in tool integration — the template ships with a weather tool example showing how to extend model capabilities with external function calls.
• Reasoning model support — the chat UI handles models that expose reasoning output, not just plain completions.
• shadcn/ui component set — the interface is assembled from shadcn/ui components styled with Tailwind CSS for a responsive layout.
• Next.js App Router — the app is built on the latest Next.js App Router conventions.
• One-click Vercel deploy — a deploy button clones the repository and provisions the DeepInfra API token integration through the Vercel Marketplace.
• Local development flow — Vercel CLI commands (vercel link, vercel env pull) pull environment variables into a .env.local file.

Who is it for?

• Developers starting an AI chat product: clone the repo and get a streaming chatbot UI with provider wiring already in place.
• Teams standardizing on DeepInfra: use the template to connect a Next.js frontend to DeepInfra models without writing provider glue code.
• Developers prototyping tool calling: adapt the weather tool example to test function-calling flows against real models.
• Vercel users: deploy the app through the Vercel Marketplace integration without manually managing API keys.

What can you do with Vercel x DeepInfra Chatbot?

• Swap model providers: change the provider configuration to route the same chat UI to a different AI SDK-compatible backend.
• Add your own tools: replace the weather example with a tool that queries your own API and observe the model invoke it.
• Self-host a chat interface: deploy to Vercel or run locally at http://localhost:3000 after installing dependencies.

How does it work?

Clone the repository and install dependencies with npm, yarn, or pnpm, then install the Vercel CLI and run vercel link followed by vercel env pull to populate .env.local. Start the dev server with npm run dev and open http://localhost:3000 to use the chatbot.

FAQ

Is Vercel x DeepInfra Chatbot free?

The template is open source and hosted on GitHub under the vercel-labs organization, so you can clone and self-host it. Using it requires a DeepInfra API token, which is connected through the Vercel Marketplace integration.

What stack does it use?

It uses Next.js with the App Router, the AI SDK by Vercel for streaming and provider abstraction, DeepInfra for inference, and shadcn/ui components styled with Tailwind CSS.

Can I use it with providers other than DeepInfra?

The AI SDK allows multiple AI providers to be used interchangeably with a few lines of code, so the chatbot layer is not locked to DeepInfra, though the deploy integration is configured for a DeepInfra API token.

Does it support tool calling?

Yes. The template includes a weather tool example that demonstrates how to wire tool integration into the chat flow.