Chat Experience
The KNVRT chat interface is designed for speed, flexibility, and absolute privacy. From model switching to deep research, every tool is at your fingertips.
Model Selection
Switch between providers mid-conversation. KNVRT maintains context even as you change the underlying intelligence, allowing you to compare answers instantly.
Deep Research
Toggle Web Search to enable RAG. The model will browse the live web to find the most up-to-date answers and provide clickable citations.
Power Features
Simultaneous Chat
Chat with multiple models at the same time. Open parallel streams with GPT-4o and Claude 3.5 to compare reasoning, coding styles, or creative outputs in real-time.
Model Switching
Switch models mid-conversation without losing context. Start a project with a fast model and swap to a high-reasoning engine like O1 for final auditing.
Incognito & Assets
Privacy doesn't limit functionality. Upload images and documents even in Incognito Mode—all assets are handled in-memory and never permanently stored.
Pro Tips for Power Users
Use the dynamic suggestion tags at the bottom of the chat to quickly steer the conversation or request refinements without typing.
KNVRT automatically detects when you need web search. Toggle the globe icon to force real-time data retrieval with full citations.
- Use Cmd/Ctrl + K to quickly open settings or toggle search.
- Compare model outputs by switching the model mid-thread and asking it to critique the previous answer.
- Enable High-Density UI for a more compact chat view during intense research.
Session Recovery
While normal chats auto-save, we recommend syncing your account to ensure your chat history is accessible across all your devices.