How much funding has Machenta raised?

Machenta has raised $40M in total funding. Their latest funding round was Seed. Key investors include Creandum, Union Square Ventures, Hoxton Ventures.

Where is Machenta headquartered?

Machenta is headquartered in London, United Kingdom. The company has a team of 11-50 people.

What does Machenta do?

Machenta operates in the Robotics / Embodied AI sector. Imperial College London spinout building a robot foundation model, trained entirely in simulation, that lets robots learn new physical tasks instantly from a single demonstration. Removes the days to weeks of real-world data collection and post-training needed to teach robots each new task. Show the robot a task once and it performs it immediately, with no further training.

When was Machenta founded?

Machenta was founded in 2025. It is based in London, United Kingdom.

How much funding has Machenta raised?

Machenta has raised $40M in total funding. Key investors include Creandum, Union Square Ventures, Hoxton Ventures.

Who is the CEO of Machenta?

Edward Johns is the CEO of Machenta. The company was founded in 2025.

Where is Machenta located?

Machenta is headquartered in London, United Kingdom.

What funding stage is Machenta at?

Machenta is currently at the Seed stage.

How many employees does Machenta have?

Machenta has approximately 11-50 employees.

Machenta

Category: Robotics / Embodied AI

Imperial College London spinout building a robot foundation model, trained entirely in simulation, that lets robots learn new physical tasks instantly from a single demonstration. Machenta was founded in 2025. The company is led by Edward Johns. Based in London, United Kingdom. Team size: 11-50. Total funding raised: $40M. Latest round: Seed. Key investors include Creandum, Union Square Ventures, Hoxton Ventures.

Founded: 2025
Headquarters: London, United Kingdom
Team size: 11-50
Total funding: $40M

Value proposition

Removes the days to weeks of real-world data collection and post-training needed to teach robots each new task. Show the robot a task once and it performs it immediately, with no further training.

Products and solutions

Vision-Demo-Action (VDA) robot foundation model for Instant Robot Learning: one-shot, in-context task learning. Demonstrated on packaging, assembly and tidying tasks across multiple embodiments, including humanoids.

Unique value

Described by the company as the first robot foundation model that has learned how to learn new tasks instantly from one demonstration, trained without a single pixel of real-world data. The company calls this 'physical vibe coding'.

Target customer

Robot operators and manufacturers, including industrial and factory automation and humanoid robot makers. The company is currently research-focused and has no near-term commercialization plans.

Industries served

Robotics, Manufacturing / Factory Automation, Logistics / Packaging, Home and Service Robotics

Technology advantage

Meta-learning through large-scale simulation data generation, so the model learns how to learn rather than memorizing individual tasks. This avoids costly real-world data collection. It builds on six years of Imperial research: self-supervised learning (2021), Implicit Graph Alignment (CoRL 2023), Instant Policy (ICLR 2025) and Learning a Thousand Tasks in a Day (Science Robotics). Every author of these papers now works at Machenta, alongside researchers from DeepMind, Dyson, Huawei and Stability AI.

How they differentiate

Competitors scale real-world data collection and post-train VLA models for each task. Skild AI's S1 does in-context learning from video demos but relies on large real-world data and has commercial scale. Machenta trains entirely in simulation and conditions on a single demonstration instead of language, so new tasks need no further training.

Main competitors

Physical Intelligence, Skild AI, FieldAI

Key partnerships

Imperial College London (spinout from the Robot Learning Lab)

Major milestones

2020: one-shot robot learning research begins at Imperial's Robot Learning Lab. 2025: Instant Policy (ICLR 2025) and Learning a Thousand Tasks in a Day (Science Robotics cover) prompt the founding of Machenta, Hoxton Ventures pre-seed investment. Oct 2026: emerges from stealth with a $40M seed co-led by Creandum and USV, the largest UK robotics seed to date.

Growth metrics

About 11 employees at the October 2026 seed round. Moving into a new office and lab in King's Cross, London.

Market positioning

Early-stage frontier physical-AI research lab taking a contrarian simulation-only approach. It competes with multibillion-dollar US robot foundation model companies. Its round is the largest robotics seed raised in the UK to date.

Geographic focus

Based in the UK/Europe and competing globally, mainly with US (California) robot foundation model companies.

Patents and IP

Builds on published Imperial College research: Implicit Graph Alignment (CoRL 2023), Instant Policy (ICLR 2025) and Learning a Thousand Tasks in a Day (Science Robotics). No public patents identified.

About Edward Johns

Professor of Robot Learning and Director of the Robot Learning Lab, Imperial College London (on faculty since 2018); began a one-shot robot learning research group in 2020; co-author of Instant Policy (ICLR 2025) and Learning a Thousand Tasks in a Day (Science Robotics cover).

Latest news about Machenta

Machenta raises $40M seed round for simulation-trained robot learning

More Robotics / Embodied AI companies

Official website: https://machenta.com