back
218 comments
> improve foundational features like the Blender Python API, which enables developers and artists alike

So they want claude to be able to talk to blender

This might actually be quite nice - the Blender Python API is currently very useful and very touchy. Lots of differences in behavior in headless mode which are hard to debug (because you can't open the GUI to see what's happening, because that changes the behavior).
There's already an MCP for it, saw a post on LinkedIn the other day about it.

Not sure if this one was the one I saw, but Google gave me this one. You could use Claude Code to build things with Blender.

https://blender-mcp.com/

We (I) need that.

"Some software" is approaching levels of complexity where, perhaps, it gets to a point where a human is barely able to even use it.

At the same time (brave new world) LLM assisted software opens up the possibility of levels of complexity we would not have considered before.

Can't wait for Gimp automation, so i can finally start using it!
Yeah, Google has MuJoCo so it seems natural to get hooks into Blender.

MuBlE: MuJoCo and Blender simulation Environment and Benchmark for Task Planning in Robot Manipulation: https://arxiv.org/abs/2503.02834

There is already a Blender MCP. It works-ish! But could be a lot better in understanding 3D space.

As an amateur this is really exciting - but not sure about folks that are real pros at this stuff.

I want claude to talk to blender, I personally hate using blender but love it's outputs
Frankly, I love the idea of an automation engine printing out tangible works. I actually build spritesheets that way! Load a bunch of individual gimp files as layers, set them offset by a given parameter, and boom, done!

Would be rad to incorporate some statistical procedurally generated designs based on my own aparatus.

What I do not want to see is this realm of LLMs hijacking decades of hard work and consideration for integration channels to more tailor towards their LLMs, not for the diligent engineer.

If they want to put their tentacles as far as they want while making products more difficult to work with innovation of a different color, they are making enemies out of, at least me.

Not sure why this is getting backlash. Just look at https://fund.blender.org. Other corporate sponsors are Google, Meta, Nvidia, Netflix, even Adidas.

This just means more support for a major OSS project.

Can you imagine going to a football match and second-guessing which are the players who look human, but skin-deep are actually androids made at a factory? This is what it feels like with music and literature right now with so much AI. There are some pockets where you still can say "that's human-made", like 3D-rendered feature films with some particular artistic direction. That, it seems, AI companies also want it to go the way of the dodo.
I just wanted to note a couple of things

1. That looks great and I hope they can make blender even better, or pay people to do that. Raise your rates for corps Blender!

2. What does this have to do with any strategic position for Claude. Their goals are

“Enterprises pay us a fortune to either

1. Sift through unbelievable volumes of data to pick out fairly easy to find nuggets

2. Tell everyone their jobs are at risk to keep them in line”

There is no other play. AGI is science fiction.

So why are they launching this? Because they don’t have a strategic core running through - it’s always been “wow what else can this do”. That’s a research project at the price of billion dollar data centres.

Unless we do achieve AGI (which we won’t) the price tag is way beyond the returns we are seeing.

I've been using Claude with OpenSCAD to generate some simple models with repetitive geometry (a set of d8 dice with braille on them for a scrabble-like game for blind children). It's really good, though often I have to send a screenshot to Claude or describe a geometry issue.

Having more native integration into Blender, which I'm already much more familiar with, will be fantastic.

Almost every AI lab I talk to that deals with 3D has built their pipelines around Blender.

This is unsurprising as a general development other than Anthropic doesn’t have a 3D model generation framework.

I don’t think this is to create MCP servers necessarily but rather to improve the blender pipeline further.

People seem not to be aware of this: https://fund.blender.org/funding-policy/

I agree that it's not a good look for Blender, but I don't think that something actually bad will come from this. (Other than maybe a negative impact to Blender's reputation.)

I'm thrilled for the world where I can drive more things with an LLM. The big limiting factor for me for little home improvement things was that I'm not very good at modeling so I have to get my wife to go do things for me. That's fine, she's happy to do it, but sometimes you kind of just have to try yourself to see what you're really looking to do.

Recently I've been using Claude Code with `build123d`[0] and it's pretty good, but my wife uses Blender so it would be cool to come up with something at least halfway decent and then have her clean it up.

0: https://wiki.roshangeorge.dev/w/Blog/2026-04-24/Modeling_Bet...

People surprised by Anthropic getting in on Blender funding obviously never saw any of the Blender/ChatGPT integrations a few years ago. This has been coming for a long time.

https://www.youtube.com/watch?v=xhN7P7ENu4g

That was ancient times in LLM terms. I've seen demos that create whole scenes in a single prompt.

There are lots of non-coding use cases for LLMs that don't burn through compute but are still useful. Anthropic is starting to catch on, and it makes sense to focus there with the compute crunch.

I prefer that outcome to them acquiring OSS projects and shutting them down once the bubble collapses.
I hope this drives the move towards what I have kept on saying is "good use" of generative ai: being able to generate animations.

It's tedious work, not the most fun thing to do especially for units like random enemy #20.

I can imagine they would be interested in creating features to generate 3d models.
There will be community backlash. And it will not be uncalled for. Sad news.
Not to be a hater or anything (I'm a hater), but I'm seeing people mention the potential of LLMs for the "grunt work" like retopo, but I can't really begin to imagine what the "correct" data representation and python api calls would even begin to look like in a training set? Would an LLM really be querying vertices in relation to one another and estimate whether their distance "sounds" like good topology?
Has anyone found a really good ai modeling and animation workflow? Ive tried a few like Meshy which work but I am not super happy with the results.
I was wondering the other day if Ai could do tedious things like retopology and figuring out effecient uv unwrapping

Also was wondering how'd it would do things like sculpting? That sounds expensive like either you send millions of polygons for the model to explore? And that ruins the context window order doesn't even fit, or your sending tons of screenshots?

I have mixed feeling about this. Guess they can need the money ... but still. Data goes to Anthropic here. It also will buy influence in some ways, I am sure about that. We could see this with rubygems.org - when shopify threatened to cut funding some months ago, suddenly chaos erupted. Money buys influence, this is easy to see how.
Mixed feelings about creative work and AI, but if it wasn't for LLMs I would have chosen a different software than Blender for my hobby-level 2D animation. Made a Blender plugin w/ Claude, and it's saving me so much time (:
This makes sense. Blender has had (non-LLM) generative features for a long time, hooking LLMs into the Python API to generate art makes sense (it's probably already done but for it to be sponsored is nice).
The press watching side of me only has questions. Why was this published by Blender and not Anthropic? What does this actually mean? That the blender team gets free claude code max subscriptions?
Any ideas why anthropic is interested in blender funding?
It's only a matter of time before parametric approaches (think LaTex for CAD, ie: via a descriptor language) enable LLMs to start 3D modeling.
The backlash is because everytime opensource projects gets involved with lots of money, everything goes to shit.
Anthropic right now internally (probably):

“We love art :P”

Will Blender start allowing Anthropic to train on your art automatically unless you opt-out?
This will be a big step up from using python to generate shitty gifs
That's already blowing up on Mastodon. Blenderartist silent for now. Won't stay like that for very long.

I wonder, if Ton was involved in that decision, or if it's only Francesco. Could turn out to be a very unlucky start into the leadership role.

I love blender. They should get all the money they need.
it would be funny if this was meant as a goodwill gesture from anthropic to counter all the recent bad press, only for it to cause even more drama
Anthropic is BLEEDING through cash as we speak while valuations soar and they have the funds to do charity?

I feel this a thinly veiled attempt at again, stealing IP.

So Claude will soon do 3D modelling?
These AI companies need to be kept away from Open Source projects. A sad day for Blender :(
Ugh. I love Blender, it's the greatest software of all time according to myself, and I absolutely hate this and I am terrified at what it implies. If they just want name recognition ok fine, but my guess is Anthropic will want changes to Blender itself and I find that totally unacceptable.

Ah well, the online artist community is unusually principled on matters like this, especially compared to here. If they start doing shady stuff it will get forked and probably spell the end of the Blender foundation, which would still be really bad of course.

Sigh. Not a happy Tuesday.

Is this going to be another app that I have to make sure I opt-out of training for?
I hope this doesn't mean enshittification of Blender.
I'm so sick of it. I'm so fucking sick of it.
It's because LLMs will soon start building real-world objects via CAD. This is the first step. Look at things like the Adam plugin for Onshape. Works great with Opus. It built a toy car for me with one prompt.
I see we're now entering the "Sam Bankman-Fried" stage of buying goodwill.