Next.js Chatbot with local Ollama, OpenAI, and DeepSeek support with Tailwind

This is a Next.js chatbot application that integrates with Ollama, the OpenAI API, and the DeepSeek API.

New version

Legacy version

Features

• Select Provider: You can select local Ollama, OpenAI, or DeepSeek.
• AI-Powered Chat: Communicates with an Ollama server to generate responses using models like deepseek-coder-v2:latest.
• Real-Time Streaming: Displays bot responses as they stream from the Ollama API.
• Markdown Formatting: Renders bot messages with Markdown support for headings, lists, code blocks, tables, bold/italic text, and links.
• Responsive UI: Clean, modern interface with user and bot message bubbles, optimized for desktop and mobile.
• Configurable Model: Allows specifying the Ollama model via a prop (e.g., llama3.2, deepseek-coder-v2:latest).
• TypeScript Support: Built with TypeScript for type safety and better developer experience.
• Customizable Styling: Uses Tailwind-inspired CSS classes for easy style customization.

Prerequisites

• Node.js: Version 18 or higher.
• Ollama: Installed and running locally (or on a specified host). See Ollama Installation.
• Git: For cloning the repository.
• npm or yarn: For managing dependencies.

It has a feedback system

Getting Started

1. Clone the Repository

git clone https://github.com/csabika98/NextJS-AI-Chatbot.git
cd NextJS-AI-Chatbot.git

2. Install Dependencies

Install the required Node.js packages:

npm install
# or
yarn install
# or
pnpm install

Key dependencies include:

• next, react, react-dom: Core Next.js and React libraries.
• react-markdown, remark-breaks, remark-gfm: For Markdown rendering.
• @next/font: For font optimization (Poppins font).

3. Set Up Ollama

1. Install Ollama: Follow the instructions at Ollama Download to install Ollama on your machine.

2. Start the Ollama Server:

   ollama serve

   This runs the server at http://localhost:11434 by default.

3. Pull a Model: Download the desired AI model (e.g., deepseek-coder-v2:latest):

   ollama pull deepseek-coder-v2:latest

   Verify available models:

   ollama list

4. Configure Environment Variables

Create a .env file in the project root to configure the providers you want to use.

These two are mandatory for Ollama:

NEXT_PUBLIC_OLLAMA_HOST=http://localhost:11434
NEXT_PUBLIC_CHATBOT_OLLAMA_MODEL_NAME=deepseek-coder-v2:latest

These two are mandatory for OpenAI. Keep OPENAI_API_KEY server-only; do not prefix it with NEXT_PUBLIC_.

NEXT_PUBLIC_OPENAI_DEFAULT_MODEL=gpt-4.1-2025-04-14
OPENAI_API_KEY=<your_openai_key>

These are required for DeepSeek. The model defaults to deepseek-v4-flash if NEXT_PUBLIC_DEEPSEEK_DEFAULT_MODEL is omitted.

NEXT_PUBLIC_DEEPSEEK_DEFAULT_MODEL=deepseek-v4-flash
DEEPSEEK_API_KEY=<your_deepseek_key>

You can combine them:

NEXT_PUBLIC_OLLAMA_HOST=http://localhost:11434
NEXT_PUBLIC_CHATBOT_OLLAMA_MODEL_NAME=deepseek-coder-v2:latest
NEXT_PUBLIC_OPENAI_DEFAULT_MODEL=gpt-4.1-2025-04-14
OPENAI_API_KEY=
NEXT_PUBLIC_DEEPSEEK_DEFAULT_MODEL=deepseek-v4-flash
DEEPSEEK_API_KEY=

If using a different host or port, update the value accordingly.

5. SYSTEM PROMPT

Feel free to edit the base prompt.

1. Go to src/app/config/systemPrompt.ts

const SYSTEM_PROMPT = `You are a helpful assistant. If i type "test" you should answer "test" only, Format all responses in Markdown, using appropriate syntax for headings, lists, code blocks, tables, and other elements where applicable. Use single backticks (\`) for inline code (e.g., \`StringBuilder\`) and triple backticks (\`\`\`) with language identifiers for code blocks (e.g., \`\`\`java\ncode\n\`\`\`).`;

export default SYSTEM_PROMPT;

6. Test the Chatbot

• Type a message in the input box and click "Send" or press Enter.
• The chatbot will respond with AI-generated text, formatted in Markdown (e.g., lists, code blocks, tables).
• Example queries:
  • "List three programming languages."
  • "Write a Python function to reverse a string."
  • "Create a table of animals and their habitats."

Style Customization

Styles are defined in app/globals.css and inline in components. To modify the appearance:

• Message Bubbles: Update userMessage and botMessage classes in ChatBox.tsx.
• Markdown Elements: Adjust the components prop in ReactMarkdown to style lists, code blocks, tables, etc.
• Global Styles: Edit app/globals.css for app-wide changes.

Example: Change the bot message background color in globals.css:

.botMessage {
  background-color: #d1d5db; /* New color */
}

Learn More

• Next.js Documentation: Learn about Next.js features and APIs.
• Ollama Documentation: Explore the Ollama API and available models.
• React Markdown: Understand Markdown rendering options.
• Vercel Font Optimization: Details on using @next/font.