Robocurve PBC

Public Benefit Corporation helping society understand the frontier of physical AI. Backed by Y Combinator.

[Initialized Capital @Initialized]

Congrats to @chooi_jeq and the @robocurve team! We’re so excited to work with you! https://t.co/GNbXjUUZnS

[Initialized Capital @Initialized]

We're thrilled to lead the $10M seed round for @robocurve, a Public Benefit Corporation independently evaluating and benchmarking how frontier AI models perform in the physical world. Frontier models are becoming increasingly capable of controlling robots, and understanding what these systems can actually do is becoming more important.

Founder and CEO @chooi_jeq is not new to this. He was previously a Research Fellow at MATS, a researcher at the UK AI Security Institute, and a top contributor to Inspect Evals, the UK government's AI eval framework.

Robocurve provides independent evidence of robotic capabilities and how quickly those capabilities are changing, helping the public understand the pace of AI progress in the physical world.

In just three months, Robocurve's research has been viewed 6M+ times, its open-source evaluation harness has been downloaded 97k+ times, and researchers from 200+ institutions have signed up to build robotics benchmarks with the team.

[Omar Rayyan @omarrayyann]

All models used the same closed-loop harness in @robocurve’s results. they get exocentric/egocentric rgb observations from the DROID setup and robot state, and output cartesian poses. history is cleared after each of the 200 episodes.

[Eric Jang @ericjang11]

@chooi_jeq @robocurve Congrats! Extremely bullish and noble mission

[Paul Hsu @paulhsu]

AI is expanding beyond software and into the physical world.

As the frontier advances, the bottleneck is shifting from capability to deployment. At the center of deployment are evaluations and benchmarks.

Above all, economic evaluations need to show how reliably and cost-effectively a system performs the work it was deployed to do.

RoboCurve is building the evaluations and benchmarks that physical AI needs to scale across any model, any embodied AI device. This scaling will soon encompass infinite tasks, jobs, and environments.

The team has already shipped Inspect Robots, the leading open-source evaluation harness for physical AI, with an extraordinary 81 releases in eight weeks. That pace of execution is rare. Speed in execution is a moat when addressing infinite possibilities.

What impresses us most is @chooi_jeq and the exceptional group of domain experts he has assembled around him. Great companies are built not only on bold ideas, but also on the depth of the people who make those ideas credible.

Congratulations to Jay and the entire @robocurve team on announcing their $10M round.

@decasonic is proud and privileged to invest this journey. I’m grateful to be included in what they are building.

[Amie @Amieaicrypto]

Congrats! One more strong player in robotics area

[Ben Hagag @BenHagag20]

No better timing! Really happy for @chooi_jeq and his team. This is exactly the kind of org we need right now

[Justin Patel @justpat]

Excited to back @robocurve!

As frontier models increasingly move into the physical world, independent evaluation becomes critical infrastructure. Robocurve is building that layer, testing what frontier AI can actually do on real robots.

Congrats to @chooi_jeq on the seed.

[Abdul @abdulalali]

we @decasonic are incredibly excited to be supporting @chooi_jeq and the wider @robocurve team. Robocurve is developing the trusted and neutral layer for physical AI benchmarks at scale.

we believe the future of the physical AI market is going to be (1) multi-model and (2) multi-interfaced. software LLMs are expanding their capabilities to the physical world (spatial reasoning from Astra) and multiple RFM providers are emerging with competitive advantages (across ICL and scale of data). the result is a fragmented market that needs a neutral and trusted benchmark that unlocks deployment of physical AI devices at scale. this benchmark is Robocurve.

@chooi_jeq's background is impressive, and their team shipped the leading open-source harness for physical AI evaluation in the form of Inspect Robots.

[Damian Player @damianplayer]

the LM Arena for robots just raised $10M.

6,000,000 views on their research in under 3 months. 97,000 downloads of the open-source eval harness.

generational run incoming.

[Notable Capital @notablecap]

Proud to back @chooi_jeq and the @robocurve team!

cc: @hanstung

[Valerio Pepe @ValerPepe]

We need more evaluators, and I'm excited to see one specialized in robotics! Excited for @chooi_jeq and @robocurve to push on this :)

[Deborah Jacob @isthatdebbiej]

Congratulations @robocurve and @chooi_jeq

We definitely need third party evaluation for frontier AI in the physical world.

[Varun Nair @_varunnair]

Congratulations @chooi_jeq and @robocurve to earning the right to solve one of the hardest problems plaguing robotics.

Evaluation of Frontier AI in the physical world!

[Decasonic @decasonic]

Physical AI is moving quickly into the real world. That makes evaluation infrastructure increasingly important.

At Decasonic, we believe the next phase of physical AI will require trusted benchmarks that can measure how models and embodied systems perform across real tasks, environments, and hardware.

@robocurve is building toward that layer.

As the ecosystem expands across models and embodiments, shared benchmarks can create the confidence needed to make deployment decisions and scale adoption.

We’re excited to support @chooi_jeq and the Robocurve team following their $10M seed.

Congratulations to the entire team 👏👏

[Mike McCormick @MikeMcCormick_]

We need third party evaluators for Ai in the physical world too

Excited to support @chooi_jeq and @robocurve!!

[Jay Chooi @chooi_jeq]

We raised a $10M seed for @Robocurve, an independent Public Benefit Corporation, to evaluate frontier AI in the physical world.

In less than 3 months, our research has been viewed 6M+ times and our evaluation harness downloaded 97k+ times. Researchers from 200+ institutions, including 19 of the world’s top 20 universities, have signed up to build benchmarks with us.

We welcome more independent evaluators. The more third-party evaluators, the better society can understand how fast robotics AI is progressing. Our evaluation harness is fully open-source, and we publish benchmarks anyone can run and verify.

Our round is led by @Initialized, with participation from @notablecap, @decasonic, @ycombinator, @HalcyonFutures, and many others.

Building robots or frontier AI? Work with us to help the world understand what your systems can do.

[Physical AI Safety Institute @paisi_ai]

Robot foundation models are shipping. What about their safety?

@drmapavone, @andrea_bajcsy, @vikassindhwani and @thomas_fel_ are speaking at the first workshop on it @ CoRL 2026, Nov 12.

$20k in travel grants - by @robocurve.

Papers due Oct 1.

https://t.co/dVPXqPN8s4

[Jay Chooi @chooi_jeq]

GPT-6 Astra consistently scores better than MolmoAct2 across all five bimanual robot tasks. In 99 of its
100 trials, Astra scored at least as high as MolmoAct2’s best trial on the same task.

[Jay Chooi @chooi_jeq]

GPT-6 Astra scores 46% vs 12% for MolmoAct2, a state-of-the-art robotics VLA, across 200 trials on five bimanual tasks. That's 3.9x higher. 🧵