Flexion

We are building the autonomy stack for humanoid robots - from command to control, from manipulation to locomotion, across any hardware and task.

[Flexion @FlexionAI]

We’ve raised $50 million in Series A funding from DST Global Partners, @nvidia (NVIDIA’s venture capital arm), @redalpine , @prosusventures , and @Moonfire_VC , following our $7.35M seed round from @frst_vc , @Moonfire_VC , and @redalpine just a few months earlier, to build the autonomy stack that makes humanoid robots adaptive, intelligent, and ready for real-world deployment at scale.

In less than a year, our team has shown that long-horizon whole-body humanoid control can scale across hardware and tasks by leveraging the power of simulation and reinforcement learning.

This funding will help us grow our team, scale our compute and robot fleets, and accelerate the commercialization of our autonomy stack with OEM partners globally.

You can find more details in the links shared in the comments.

[Flexion @FlexionAI]

Read the full technical breakdown here: https://t.co/iIKaLlQOAu

[Flexion @FlexionAI]

"A parcel with snacks has been delivered for Flexion. Retrieve it using the stairs and come up using the elevator. Then unpack it and place the items into the empty drawer on the shelf in the snack area."

One instruction. No human operator. Everything that follows is autonomous.

Today we're introducing Reflect v1.0, our robotics intelligence platform for long-horizon work. From a single natural-language command, the robot understands the task, navigates a multi-floor building, calls elevators, handles doors, uses tools to unpack a box, and puts the items away. The biggest shift in v1.0 is that we use reinforcement learning across every layer, from low-level control to high-level reasoning.

Long-horizon autonomy is unforgiving. The robot must recover on its own when things don't go to plan because in the real world, they never do. Combining reasoning, perception, physical execution and runtime robustness into a single mission-capable system is the foundation required to solve humanoid autonomy.

Our team is just getting started.

#HumanoidRobots #Flexion

[Flexion @FlexionAI]

To accomplish that, our stack ran end to end with no human involvement, leveraging our command layer to break down a bigger task “go over the stairs, grab the cardboard box on the floor, set it down on the far table, and head back to your starting point” into actions that our motion and control layers could execute, chained together in a single fully autonomous sequence.

300 demos, with above 95% success rate. That's a good start, but as we keep collecting data on the robot operating in live environments, we can use it to improve our intelligence layer and go from 95%+ to 99.99% success rate. That's what we're building at Flexion: a truly autonomous system that handles tasks end to end and scales to real-world deployment.

More coming soon!

[Flexion @FlexionAI]

We ran 300 fully autonomous live demonstrations over 3 days at ICRA 2026.

The task: a humanoid navigating stairs, picking up a box from the floor and placing it on a table. Simple to describe, but hard to execute reliably when your robot is making every decision on its own at a conference with new surroundings and a crowd watching live.

This is just a glimpse. We've been pushing our stack much further and we'll be sharing more very soon.

More information in the thread.

#HumanoidRobots #ICRA2026 #Flexion

[Flexion @FlexionAI]

Flexion is present at ICRA, stand 89. Bring your questions. Meet our team. And if you're exploring what's next in your career, connect with us.

Stand 89, Vienna.
June 1st - 4th

#HumanoidRobots #Flexion

[Flexion @FlexionAI]

Last week in Zürich, we co-hosted a panel with @foxglove at #ActuateFieldSessions around an honest challenge: Why general purpose robot learning hasn’t had its breakthrough moment yet?

The answer isn't more data, it's the right data: touch and real-world scenarios that perception alone can't capture.

The data gap is real, but we’re closing it.

Great conversation with @IlirAliu_ , @HoellerDavid, @Klajd_Lika, Mayank Mittal, @arbwes and the Foxglove team.

#HumanoidRobots #Flexion

[Flexion @FlexionAI]

Read the full article here: https://t.co/TR8z3OrSzx

[Flexion @FlexionAI]

Robots are getting smarter and the way we build them is changing fast.

Simulation + RL can now train a robot in days, not years.

Our CEO @rdn_nikita sat down with Evan O'Donnell for The Times Blog to unpack the process behind teaching robots how to move.

Full article in the comments.

#HumanoidRobots #Flexion

[Flexion @FlexionAI]

Sign up for the session here: https://t.co/bhvMtnbNP9

[Flexion @FlexionAI]

Robotics is talking about the data gap. How do we solve it?

Co-hosting #ActuateFieldSessions alongside @foxglove - with @HoellerDavid, @Klajd_Lika, Mayank Mittal, and @arbwes on the panel, moderated by @IlirAliu_

May 6th | Zürich

Join us - link in the comments.

#HumanoidRobots #Flexion

[Flexion @FlexionAI]

@MartinKiefel Add this session to your VENTURE agenda: https://t.co/WdFuaViZwf

[Flexion @FlexionAI]

Physical intelligence isn't a headline for the future. It's what we build and deploy today.

Our Co-Founder and Perception Lead, Julian Nubert, is having that conversation with @MartinKiefel, Sören Heß, and Leonard Schenk at #VentureSprind2026 in Berlin - diving into how physical intelligence is redefining work and the future of European industry.

Key players. Bold approaches. Tomorrow's economy.

Session link in the comments.

#PhysicalIntelligence #VentureSprind2026

[Flexion @FlexionAI]

What does it take to build the brain of a humanoid robot?

Great having @andreasklinger at the office to see it firsthand — from robots learning to move through reinforcement learning, to the bigger vision of one AI platform for any humanoid hardware.

[Andreas Klinger 🦾 @andreasklinger]

Twitter boi, do you want to see a humanoid AI lab from the inside?

Well I do ;)

Join me to visit @FlexionRobotics
A world-leading lab building the AI brains for humanoids – in Zurich, one of the best hubs for Physical AI.

Their goal? Build the Android of robotics.
One operating system that works on any humanoid hardware.

It is crazy what their robots can do. We pushed a robot on the stairs. It didn't care. They left one in a rainy forest it was never trained on. Chilled and hiked there.

And because they let robots train on their own, from zero through reinforcement learning, the robots do odd inhuman things. Eg they stand up from the ground in ways no human ever would.

If their humanoid brain works, this will be industry defining. This is why Flexion is one of Europe's Most Ambitious Startups. 🇪🇺🔥

0:00 Intro
0:38 Inside a Lab with 14 Humanoids
2:17 What’s Actually Inside a Humanoid?
4:43 The Robot Navigates on Its Own
6:31 Why Grasping Is So Hard
7:01 Stairs, Pushes, and Balance Tests
7:55 Training 4,000 Robots in Parallel
10:50 Teleoperation vs Reinforcement Learning
14:35 The Case for Humanoids
15:55 Why Europe Could Win Robotics

Please like & RT 🙏

[Flexion @FlexionAI]

While others rely on teleoperation, we are betting on simulation at scale. Here is how Flexion is building the data flywheel differently. 👇

[Forward Future @ForwardFuture]

"Teleoperation in robotics is very popular right now."

"We’re intentionally avoiding it."

@rdn_nikita CEO & Co-founder @FlexionRobotics on how they’re training robots at scale:

“We’re betting heavily on simulation and reinforcement learning.”

“No motion-capture suits. No VR headsets. No armies of people piloting robots.”

“Instead, we train as much as possible in simulation.”

“If a new robot comes in, we load its URDF, retrain in the simulator, and deploy a new neural network.”

“So when we train one robot on a task, we’re effectively training dozens or hundreds of embodiments at once.”

“That’s the flywheel.”

[Flexion @FlexionAI]

“My hot take is this: there isn’t a single humanoid robot today that truly generates value.

Some robots can do something close to the intended task. But not the actual task. And if a robot needs human handlers to clean up after it, you’re not creating value. In many cases, the value is arguably negative.

We’ll fix that.”

Our Co-Founder and CEO, @rdn_nikita, joined @twimlai to share a grounded view on where robotics stands at the beginning of 2026 and how the next few years will reshape the industry.

🎧 Listen to the full conversation below.

[The TWIML AI Podcast @twimlai]

Today, we're joined by @rdn_nikita, co-founder and CEO of @FlexionRobotics to discuss the gap between current robotic capabilities and what’s required to deploy fully autonomous robots in the real world. Nikita explains how reinforcement learning and simulation have driven rapid progress in robot locomotion—and why locomotion is still far from “solved.” We dig into the sim2real gap, and how adding visual inputs introduces noise and significantly complicates sim-to-real transfer. We also explore the debate between end-to-end models and modular approaches, and why separating locomotion, planning, and semantics remains a pragmatic approach today. Nikita also introduces the concept of "real-to-sim", which uses real-world data to refine simulation parameters for higher fidelity training, discusses how reinforcement learning, imitation learning, and teleoperation data are combined to train robust policies for both quadruped and humanoid robots, and introduces Flexion's hierarchical approach that utilizes pre-trained Vision-Language Models (VLMs) for high-level task orchestration with Vision-Language-Action (VLA) models and low-level whole-body trackers. Finally, Nikita shares the behind-the-scenes in humanoid robot demos, his take on reinforcement learning in simulation versus the real world, the nuances of reward tuning, and offers practical advice for researchers and practitioners looking to get started in robotics today.

🗒️ For the full list of resources for this episode, visit the show notes page: https://t.co/NQu3VglxjA.

📖 CHAPTERS
===============================
00:00 - Introduction
04:07 - Is robot locomotion solved?
06:04 - Sim-to-real gap
08:58 - Adding semantics to policies
09:42 - Modular vs end-to-end architectures
10:29 - Planner model
12:21 - Adapting RL techniques from quadrupeds to humanoids
15:39 - Behind robot demos
18:09 - Humanoid robots in home environments
22:03 - Training approach
23:56 - VLA models
27:59 - Closing the sim-to-real gap
32:55 - Task orchestration using VLMs
36:38 - Tool use
38:10 - Model hierarchy
43:37 - Simulator versus simulation environment
44:57 - Combining imitation learning and reinforcement learning
46:42 - RL in real world versus RL in simulation
52:58 - Reward tuning and value functions in robotics
56:38 - Predictions
1:00:10 - Humanoids, quadropeds, and wheeled platforms
1:02:45 - Advice, recommended robot kits, and community pla

[Flexion @FlexionAI]

Thank you for an amazing 2025. Happy holidays, and see you in 2026!

[Flexion @FlexionAI]

Full technical breakdown here:
https://t.co/klxQk983Ox

[Flexion @FlexionAI]

In our latest video, our agent tidies up our office fully autonomously, starting from a simple user prompt.
No scripts. No pre-computed trajectories.

Continues in the thread.

[Flexion @FlexionAI]

On top of this foundation runs our VLM-based agent: it perceives and interprets the scene, reasons about the task, and plans an appropriate strategy.

[Flexion @FlexionAI]

At the heart of the stack is our SOTA perceptive rough-terrain locomotion policy. It is trained end-to-end to handle the complexity of the real world, and deployed sim-to-real.