Ask any question about Blender, Python, Blenders Python module bpy or really anything else and receive context-aware responses. Whether you're a beginner or an expert, Chat Companion is your personal assistant for Blender. The AI usually gives you a detailed answer including python code and step by step instructions, depending on the context and task you give it.

The addon will automatically format the answer if there is any code or step-by-step instructions in it. Code can be executed immediately, fixed and copied. In instruction lists you can easily just click on one of the steps to ask for more information about it. Please keep in mind that the AI is not perfect. It sometimes generates false answers, error prone code or can have a long response time when the OpenAI servers are under heavy load.

Each prompt and answer are stored in a history (the history will stay in a blend-file when you save and reopen it). You can mark each history item as a favorite, rename them, delete them and clear the whole history while preserving favorites.

Execute directly from the answer, no copy-pasting necessary. And since the AI is not always perfect, you can also directly ask for a fix if there is an error in the script.

You can copy the code to your clipboard, a new text file or your current cursor location in a text file in Blender.

In the Blender Text Editor, right-click on one or multiple comment lines and ask for a code completion. The generated code will automatically be inserted into the text editor.

You can right-click on a lot of settings in the Blender User Interface to ask the AI about it. This will give you more information and also a python code example.

If you encounter any problems or errors, please contact me!

The addon uses the OpenAI API, so please go to https://platform.openai.com/account/api-keys to sign up and get your API key.

In order to use GPT-4, please enroll into the waitlist at https://openai.com/waitlist/gpt-4-api. It took only one day for me to be accepted.

When you use the addon, you pay OpenAI (not me) for the tokens you use (https://openai.com/pricing). For all my testing during development so far I payed $1 in total, so it is very affordable. The Addon shows you the tokens for each prompt and for the current session. At the moment the prices for ChatGPT are $0.002 / 1000 tokens and GPT-4 costs $0.03 for prompt and $0.06 for completion / 1000 tokens. 

Make sure to uninstall the current version before installing the new one, and restart Blender.

Version 1.2.7 - 16. May 2023

• Streaming now works on all Blender Versions
• Better Dependency Installation
• Some bug fixes and minor improvments

Version 1.2.1 - 24. April 2023

• Fixed a bug in Code Completion
  

Version 1.2 - 17. April 2023

• Added Code Completion [beta]
• Added the ability to ask for error fixes if a script from the AI throws an error
• Added a history of all prompt/answers, ability to set favorites, rename them, delete and clear them, also they are saved with the blend file
• If the answer includes step by step instructions, each step can be explicitly asked for with one click
• Answers are now streamed word by word, can be changed in addon preferences
• More places where you can right click on and ask the AI about in the Blender user interface
• Many smaller improvements and bug fixes

Version 1.1 - 01. April 2023

• Support for Blender Version 3.3.5 LTS, 3.4 and 3.5
• Asking the AI about a setting with the right-click context menu gives a more detailed and precise answer
• More descriptive API connection error display and better error handling (connection timeouts, heavy OpenAI server load, ...)
• Ability to change the timeout of the connection to the API

Version 1.0 - 19. March 2023

• Initial release