DeepSeek Blender Addon - Python Scripting AI | 2025

Experimental AI-powered code assistant for Blender's Text Editor using DeepSeek API.
Features:
AI-driven code completion
Automatic error fixing
Customizable prompts
⚠️ Experimental addon - use with caution
GitHub: github.com/jer...
Feedback and suggestions welcome!
Tags:
#deepseek #blender3d #blender #python #bpy

[@rejdevx]

To connect Deepseek with the Blender addon, you need to set up an API key. To do this, please visit the Deepseek Platform website at: platform.deepseek.com/ From there, you can manage your API keys. Once the website is back online (it's currently under maintenance), simply copy your API key and configure it in the addon's preferences within Blender.

[@pigcheong3283]

I have API already, and amend in blender addon, error still and hanging and hanging, I need to force quit my blender

[@rejdevx]

Try updating the addon, I made some adjustments in the latest version: github.com/jer-nc/blender_deepseek_ai/releases/tag/1.0.3 If the issue persists, please check the console messages in Blender: 1. Go to Blender > Window > Toggle System Console. 2. Try to reproduce the error. 3. Copy and paste the console messages here so I can check what's going on. Important: After setting the API key, don’t forget to save your preferences in Blender. If you don’t, the API key won’t be set correctly, and the API request will fail. Go to Blender > Edit > Preferences > Left Bottom (Hamburger Menu) > Save Preferences. Also, you could try using other APIs in case DeepSeek is overloaded. I’ve tested Qwen (qwen-max-2025-01-25) and OpenAI (gpt-4o-mini), and they should work too. For OpenAI, make sure to set your API key in the system environment variables: `OPENAI_API_KEY=` You can follow OpenAI’s quickstart guide here: platform.openai.com/docs/quickstart?desktop-os=windows Let me know if this helps!

[@pigcheong3283]

@@rejdevx this time is HTTP error 401

[@rejdevx]

@@pigcheong3283 Please check api-docs.deepseek.com/quick_start/error_codes and try saving your add-on preferences before submitting requests.

[@MusclesLikePapi]

bro music choice was sick for a tb thank you

[@tyagagerson]

we got Deepseek with Blender Bridge before gta 6

[@joshiifive]

That's crazy

[@inteligenciafutura]

It's promising, I was more or less thinking of creating something like that, but I still can't find enough information to create addons.

[@ncsmusic34]

So does it process everything on their servers or locally on our PCs ?

[@miwermachado4315]

Wait you can actually create models ? like i can write a prompt saying to deepseek create a dragon ?

[@rejdevx]

Not really, it's more for assisting with Python scripting rather than generating meshes.

[@miwermachado4315]

@@rejdevx Oh okay, thanks

[@alexsandropercy]

Any way to run it using a local LLM like using Ollama?

[@goosesuka]

Ask for a rabbit

[@goosesuka]

wow she make cuuube

[@bhushankapagate]

How can I get this?

[@rejdevx]

github.com/jer-nc/blender_deepseek_ai/releases , latest version (1.0.2)

[@Gachims]

So it can create model for me now?

[@Funblend-zd6dt]

Yes... a cube.

[@pigcheong3283]

how to connect deepseek data

[@rejdevx]

To connect Deepseek data in the Blender addon, you need to set up an API key. To do this, go to the Deepseek Platform website at: platform.deepseek.com/ From there, you can manage your API keys. Once the website is back online (it is currently under maintenance), you can copy the API key and configure it in the addon's preferences within Blender.

[@satishvfx2995]

401 error

[@rejdevx]

Please check api-docs.deepseek.com/quick_start/error_codes and try saving your add-on preferences before submitting requests.