🛠️ Available Features
Dynamic Model Selection
All models are loaded live from the Panda-Core API and grouped by capability — Text, Code, Image, Voice, and more as they are added.
Real-Time Streaming
Server-Sent Events stream tokens as they are generated, giving you instant feedback just like ChatGPT or Claude.
Parameter Control
Adjust Temperature and Max Tokens per session. Add a system prompt to steer model behaviour precisely.
Persistent History
Conversations are saved locally in your browser. Resume, rename, or delete chats at any time from the sidebar.
Model Badges & Features
Each model shows its capabilities at a glance — badges for size, specialization, and features like code generation or image understanding.
Model Details
Expand any model card to see its description, features, aliases, and technical details — no more guessing what each model does.
AI Chat FAQ
Which models are available?
All models registered in the Panda-Core AI Gateway are loaded dynamically and grouped by capability. Each model shows its description, badges, and features — just click the expand arrow to see details.
What do the badges mean?
Badges indicate model characteristics: parameter size (120B, 26B), specialization (code, text), availability (free), or status (beta, new). Hover over a badge for more context.
Is my API key safe?
Yes. The OpenRouter API key never leaves the server. You authenticate with a Panda-Core key which the gateway validates before proxying the request.
Can I switch models mid-conversation?
Absolutely. Select any model from the sidebar at any time. The full conversation history is sent with each new message so the new model has complete context.
Where is my chat history stored?
Chat history is stored exclusively in your browser's localStorage. Nothing is sent to the Panda-Core servers beyond the actual API requests.
What does the system prompt do?
The system prompt is sent as the first message with the role "system". Use it to give the AI a persona, restrict its scope, or provide background context.
What are model features?
Features describe what a model can do — for example "code", "streaming", "image-understanding", or "function-calling". Expand any model card to see its full feature list.
Will more model categories be added?
Yes. Image generation, voice/TTS, and video models are planned. The model selector automatically groups them by capability once registered in the gateway.