TypingMind setup: API keys, local storage and agents

This beginner tutorial walks through TypingMind's workspace setup and chat features. It starts with choosing a license plan and receiving a license key, then explains the storage choice. Cloud Sync uses a TypingMind cloud account for backup and synchronization; Local Only keeps data on the device. That storage option does not establish that model inference runs locally or that the app works offline. The demonstrated model connection requires a provider API key.

The setup continues in the Models tab, where users enter API keys, enable models and select one in the chat interface. The speaker also demonstrates internet research and uploading reports to generate insights and charts, with previews in the artifacts window.

For chat organization, the tutorial covers folders, pinned conversations and deletion from the sidebar. Folders can act as projects with their own instructions, models and training data. The AI agent section includes prebuilt and custom agents configured with instructions, with email writing as the example. Users can save reusable instructions in the Prompt Library.

The final sections introduce plugins and MCP connections, including a Web App Builder example. The knowledge base accepts internal data through sources such as Notion, GitHub, Google Drive and direct uploads. The speaker shows enabling it in chat so the AI can retrieve information relevant to a question.