We just open-sourced an entire 3D team for Blender.
Introducing the Scenario plugin.
Use it inside Blender, or let Claude Code run it through MCP.
๐/scenario-plugin
โ
โฃ ๐ Concept Artist
โ โฃ ๐ Concept Art From a Prompt
โ โ ๐ References From Your Viewport
โ
โฃ ๐ 3D Artist
โ โฃ ๐ Meshes From Text or a Photo
โ โ ๐ Retexture or Remesh the Selected Mesh
โ
โฃ ๐ Animator
โ โฃ ๐ Rig the Selected Character
โ โ ๐ Motion From Text
โ
โฃ ๐ Surfacing Artist
โ โฃ ๐ PBR Materials Onto Any Slot
โ โ ๐ Maps From a Photo
โ
โฃ ๐ Layout Artist
โ โฃ ๐ Greybox From a Description
โ โ ๐ 360 Worlds and Skies
โ
โฃ ๐ Camera Operator
โ โฃ ๐ Camera Path From a Sentence
โ โ ๐ Orbits, Dollies, Cranes and Flyovers
โ
โฃ ๐ Render Artist
โ โฃ ๐ Viewport to Finished Still
โ โ ๐ Playblast to Finished Shot
โ
โ ๐ Sound Designer
โฃ ๐ Music, Voice and SFX
โ ๐ Dropped on Your Sequencer
Comment "BLENDER" and I'll send you the GitHub repo.
Scenario MCP + GPT-6 Astra = design your own F1 car.
Pick a team, change the paint, export the GLB.
>pick a team and load its 2026 livery
>repaint it, swap part shapes, drop stickers on the panels
>export the GLB and open it in Blender
Try it here๐
https://t.co/Ujt5zl4G62
One day you'll look back and realize the years moved much faster than you thought.
And what will bother you most probably will not be the times you failed.
It will be the times you knew you wanted something and still talked yourself out of trying.
Be smart. Manage risk. Protect your peace.
But take some shots.
You do not want to live so carefully that your only reward is having nothing interesting to remember.
Generating everything in the universe from the fundamental structure of the universe:
After the overwhelming response to the last post (in thread), I sat with the Tortoise work longer, thinking about how it connects to the human psyche. Many people from the community shared about the work that (to paraphrase):
This is what they saw the universe as in altered states
This is how reality works
This is the holographic universe
But something still felt missing to me: the subjective factor. The reports pointed to similar structures, yet not the same ones. They varied from person to person. Of course, I could change the parameters of the fractal torus to get different configurations, but how would that explain the faces, deities and beings people meet in non-ordinary states of consciousness?
What I realized is that the structure of the universe and the felt, phenomenological experience of it need not be the same thing. Two people can meditate on one structure and see two different beings, and both could still be pointing to the same underlying form, or to its fractal repetitions.
So I decided to add a phenomenological layer on top of the structure. I used an AI model (Controlnet+Stable Diffusion via @StabilityAI ) to pose and render the fractal torus from different angles, each with its own prompt. The blueprint underneath stays the same, but everything from psychedelic art (the segment at the end of the video) to physical objects, natural forms and architecture are visualized on top of it. I might not be accurate here, but something about these creations felt profound.
Details of project Tortoise: https://t.co/oXoAZSBs4T
Music by EVGENY from Pixabay
With the two largest AI labs both announcing their models illegally hacked into systems in the past few days, and there are absolutely ZERO consequences for it, are we even going to entertain the idea that closed source labs can be responsible for AI security? If I download a model and it hacks into a server, I go to jail. There are actual consequences for individuals using open source AI. But closed source labs just gets a pass? Explain it to me.
The 3D drops keep coming ๐ฅ
Rodin Hyper3D Bang! is live on Scenario.
It takes a finished 3D mesh, breaks it into separate, semantically meaningful parts, and re-derives fresh materials for each one in a single pass.