Introducing Hedra Model Factory: an autonomous inference and training platform for building models for the real world. The recursive self-improvement loop behind frontier research has been locked inside a few labs. We're opening it to teams in robotics, manufacturing, biology and defense. Our researchers come from Stanford, NVIDIA and... Meta Superintelligence Labs. Now we work alongside your team to post-train and optimize your models.show more

We're expanding to full-service research support. Start with your data and simulation environments. We help with the entire model development lifecycle, from your first RL environment to every subsequent iteration. Every team gets forward-deployed researchers who work alongside you at each step. Here's what our researchers have been working on ↓

→ Post-training world models for physical consistency. One researcher retrained H3 across data, simulation and the hero run to beat leading frontier models on Google DeepMind's Physics-IQ (34.9 vs 30.3) → We discovered a new distillation technique that runs world models in real time without losing physical consistency, now powering our synthetic data pipeline for customers → Autonomous evals that catch defects in training data. We optimized a VLA to dramatically cut captioning and QC costs → Building infrastructure to scale past 100,000 simultaneous RL environments and simulators on our new CPU/GPU sandboxes We're publishing the research as we go, starting this week.

We're building a factory that gets faster with every model it produces, and adding thousands of GPUs to power it. Talk to our team today →

ig now's a good time to say: - if you need compute, we have compute, B300s, 200s, H100s, H200s, AMD MI355X - if you want to post-train OSS models for specialised use cases (eg. driving tractors, analysing stem cells, etc) we'll do it for you - we have GPU sandboxes dm me!!

Robotics is going to change the world. There’s a lot of untapped potential, especially as the scope of roles improves.

potential goes brr

it's actually very impressive and refreshing to see the products that reach outside the digital on the timeline 💪 what's the plan for the future?

Congrats to @mjlbach and the Hedra team!!

@mjlbach thanks Andrew!!

physical AI is the future, and whoever delivers the end to end infrastructure optimized for pixels and atoms vs. text is going to win big.

indeed. new era of models

This is super interesting

congrats 👏

This sounds like a service, right? Based on what I understand, this is not self-serve, correct?

yes its a service! we'll work with you

34.9 vs 30.3 against deepmind is no joke

Recursive self-improvement is for sure the key to actually getting to the next level! How close are we at this point to AI just continuously getting smarter, truly on its own?

we're already seeing it internally

Does it have an end or will it keep going?

Bringing the self-improvement loop to robotics and real-world data is a big step. Congrats on the launch!

congrats this is insane

thx

So excited for this! So much demand for models for the real world 🏭

accelerate

congrats!

thanks sophia!

fully autonomous looped inference and training platform is actually crazy

crazier things coming

This is so freaking cool team!!!

hell yea

Great video team!

lfg this is sick

right back at ya

banger launch :)

bang bang!!

🔥🔥🔥

🔥🫡

the stuff only the big ai labs had is now open to everyone. hedra cooking

we cooked and we ate

congrats hedra team!!

ty liv!!

Congrats on the launch!

thank you David!

I wish I had more friends IRL to discuss this with

I am so bullish on robotics long-term. I think it's going to change how a lot of humanity works for the better. But one of the biggest shortcomings right now is exactly this. The lack of the ability to adapt. Great idea with Model Factory. This is the innovation we need in the space!

this changes everything 🔥

everything's changing

massive congrats!!!

yay thanks andrew

This is the next hot space in AI

we're only getting started

this is super cool!

🔥

Amazing work Hedra team

thanks!

nice, congrats and good luck

pow pow

Could you outline a bit what a recursive self-improvement loop looks like for robotics, for instance? I mean it seems clear how it works with LLMs, where they become smarter and thus improve even faster from there. But it is not clear how this positive feedback loop could work for other types of models that are not building themselves? To me it looks more like an adaptive research strategy that is continuously finding out what the best hypothesis is to test to obtain the best knowledge for further improvement. That is still impressive though!

massive day for the Hedra team congrats y’all!

what does the testing and validation process look like before a custom-trained model is ready for actual production use?