Conversa
A full-stack MERN (MongoDB, Express.js, React.js, Node.js, Socket.IO) chatting application 💬 with friends having features like trending social chatting apps additionally with a Personal Chatbot 🤖
Last verified:
What is Conversa?
Conversa (conversa-chatting.netlify.app) is a streaming AI chat interface demo built using Netlify's AI Gateway prompt template. It provides a chatbot widget that allows users to interact with AI models including Claude, GPT, and Gemini without needing to configure API keys manually. The gateway handles authentication, rate limiting, and monitoring automatically.
Key features include a floating chat widget typically placed in the bottom-right corner of the page, message history showing both user and assistant messages, text input with send button, keyboard shortcuts (Enter to send, Shift+Enter for newlines), and progressive streaming where responses appear token-by-token. The backend uses a Netlify Function that accepts POST requests with messages in OpenAI chat format and streams responses back with Content-Type: text/event-stream.
This tool is intended for developers and website owners who want to quickly add AI chat functionality to their sites. It's particularly useful for those who want to experiment with AI chatbots without managing API credentials, for demo purposes, or as a starting point for building custom AI chat interfaces. The system prompt can be customized to scope the assistant's behavior to specific purposes like customer support, product recommendations, or general Q&A.
Conversa pricing
Pricing model: Freemium
Free to deploy on Netlify. The Netlify AI Gateway handles AI provider access without requiring separate API keys. Users may need their own AI provider accounts depending on configuration. Netlify's free tier includes the necessary functions and edge capabilities. No per-message fees from Netlify itself, but AI provider costs may apply depending on the gateway configuration and usage.
Conversa pros
- No API keys required - Netlify AI Gateway handles authentication
- Access to multiple AI models (Claude, GPT, Gemini) in one interface
- Automatic rate limiting and monitoring built-in
- Streaming responses appear token-by-token for better UX
- Simple backend implementation with Netlify Functions
- Floating widget placement in bottom-right corner by default
- Keyboard shortcuts (Enter send, Shift+Enter newline)
- Conversation history limited to last 20 messages to control tokens
- Easy system prompt customization for different use cases
- Matches existing site design automatically
- Zero setup required for AI provider credentials
- Works with any web framework
- Open source implementation available on Netlify
- Free to deploy on Netlify
- Production-ready streaming implementation
Conversa cons
- Requires Netlify platform to deploy
- Limited to last 20 messages of conversation history
- Default widget placement may not suit all designs
- Requires provider SDK installation (openai, anthropic, or google)
- System prompt scope must be manually configured
- No built-in user authentication or privacy controls
- Limited customization without code modifications
- Dependent on Netlify AI Gateway availability
Frequently asked questions about Conversa
Do I need an API key to use this chatbot?
No API keys are required. Netlify's AI Gateway automatically handles authentication, and the SDK picks up the gateway URL via environment variables. You can access Claude, GPT, or Gemini without manually configuring credentials.
Which AI models are supported?
The chatbot supports Claude (Anthropic), GPT (OpenAI), and Gemini (Google). The default provider is OpenAI/GPT, but you can configure which provider to use through the PROVIDER variable.
How do I customize what the chatbot does?
You customize the chatbot's purpose through the CHAT_PURPOSE variable, which becomes the system prompt that guides the AI's behavior. This constant is clearly marked in the code for easy editing.
Where does the chat widget appear on the page?
By default, the chat widget is placed as a floating widget in the bottom-right corner of the page. You can change this placement using the PLACEMENT variable to match your site's design needs.
How does message streaming work?
The backend Netlify Function streams the AI response back with Content-Type: text/event-stream. The frontend progressively renders the response token-by-token, so text appears gradually rather than all at once.
Can I control how many messages are remembered?
Yes, the implementation limits conversation history to the last 20 messages by default. This is done to control token usage and can be adjusted in the backend code if needed.
What keyboard shortcuts does the chat support?
You can press Enter to send a message and Shift+Enter to create a new line in the text input. This is standard behavior for chat interfaces.
Do I need to install any dependencies?
Yes, you need to install the appropriate provider SDK as a dependency: openai for OpenAI, @anthropic-ai/sdk for Claude, or @google/generative-ai for Gemini, depending on which provider you choose.
Can I use this for customer support?
Yes, you can configure CHAT_PURPOSE to scope the assistant for customer support tasks. The system prompt will guide the AI to behave appropriately for handling customer inquiries.
How do I deploy this to my own site?
Deploy using Netlify with the add-ai-chat-to-your-site prompt template. The implementation includes a Netlify Function at netlify/functions/chat for the backend and a Chat UI component for the frontend that matches your site's design.