AI Chat

PandaNpc provides a powerful and intuitive AI chat experience with support for multiple models and advanced features.

Streaming Responses

All AI responses are streamed in real-time, so you can see the output as it's being generated. This provides a more interactive and responsive experience compared to waiting for the full response.

Markdown Rendering

PandaNpc features a high-performance Markdown rendering engine built with Rust WASM:

• Code Blocks — Syntax highlighting for 100+ programming languages
• Tables — Full GFM table support
• Math — LaTeX math rendering with KaTeX
• Images — Inline image display
• Links — Clickable links with preview

Performance

The rendering engine uses a Web Worker for background processing, ensuring the UI remains smooth even when rendering large messages. Key optimizations include:

• Virtual scrolling — Only visible messages are rendered in the DOM
• DOM pooling — Reuses DOM elements to minimize memory usage
• Rust WASM parser — 10-100x faster than JavaScript Markdown parsers

Image Support

Send images along with your messages for vision-capable models:

1. Click the image button in the input area
2. Select or paste an image
3. Add your text prompt
4. The AI will analyze and respond to your image

Supported by: GPT-5, Claude Opus/Sonnet, Gemini Pro

Conversation Management

• Multiple Conversations — Keep multiple chat sessions organized
• Conversation History — All chats are saved locally
• Search — Find messages across all conversations
• Export — Export conversations as Markdown or JSON

Thinking Mode

For reasoning-capable models (GPT o-series, DeepSeek R1), PandaNpc displays the model's thinking process in a collapsible section, allowing you to see how the AI arrives at its answer.

Message Actions

Right-click any message to access:

• Copy — Copy message content
• Edit — Edit and resend a message
• Regenerate — Get a new response from the AI
• Delete — Remove a message from the conversation