Skip to content

Chatting with models

New chat and chat history

By clicking "New Chat" in the left upper corner of the application, you can start a new conversation with BeeChat. Below this button, you see your history of conversations. You can search for chats and organize them into collections.

Model selection

By clicking the model name, you can switch to or add another model. Currently, users get access to additional models on request only.

Code execution and interpretation

You can activate code execution by pressing the corresponding button below the main text input bar. If activated, you can ask BeeChat to generate python code and execute it directly in the browser. You can find more information here.

Interactive artifacts

When you prompt BeeChat to generate certain content types (e.g., SVGs, HTML files, JS diagram rendering libraries like D3 and mermaid), an artifact is generated and displayed on the right side of the screen. You can then use prompting to iteratively refine the artifact. If you accidently close the artifact window, you can reopen it in the overflow menu in the upper left corner (...). More information here.

Prompt suggestions

Below the main text input bar, you find a scrollable list of prompt templates. After having chosen a template, you can quickly skip through the placholders with the tab key.

You can activate web search by pressing the corresponding button below the main text input bar. By doing so, the model gets access to the internet and might be able to provide better results. However, be advised to always double-check the sources.

Voice recording

Press the microphone button to record a voice message which is then sent transcribed and sent to the model. We do not store any recordings. Transcription is performed locally.

Call

Using the call functions, you can have an interactive conversation with your chosen model. However, we currently do not actively support and optimize this feature.