Axis Robotics

• Axis V2

  Axis

• 2026-07-27 - Seed - $12M - Axis Robotics raises $12M seed for physical-AI data infrastructure

  Axis Robotics completed a $12 million seed round led by Hack VC to scale its distributed simulation, data-collection and online post-training platform. - source: theblock.co

• 10K Ventures - Participant - VC - 1 funding event - first seen 2026-07-27
• Hack VC - Lead - Angel / Individual - 1 funding event - first seen 2026-07-27
• Nomad Capital - Participant - VC - 1 funding event - first seen 2026-07-27
• Pi Network Ventures - Participant - VC - 1 funding event - first seen 2026-07-27

• Chris Feng - Co-founder & CEO - Founder and current chief executive

• Axis Robotics releases Axis Sim Dataset V1

  2026-09-04 - Research Publication - The public release includes the dataset, training code and benchmarks for a simulated Franka Research 3 arm. Continual pretraining on V1 increased π0.5 success on LIBERO-Plus from 83.9% to 88.8%. The trajectories cover pick-and-place, stacking, pouring, articulated-object manipulation and tool use collected through Axis Hub.

  Source
• Axis Robotics publishes the AXIS community-driven robot-learning data engine

  2026-07-23 - Research Publication - The released AXIS dataset contains 207 manipulation tasks and more than 50,000 trajectories. The system combines browser-based teleoperation with automated task generation, success checking, trajectory cleaning and visual and physics-based augmentation. Continual pretraining improved the reported overall success rate of π0.5 by 5.8% and outperformed the RoboCasa365-pretrained comparison by 37.3%.

  Source
• Axis Task 74 sample dataset – long-horizon manipulation (kettle pouring)

  2026-04-29 - Research Publication - Sample dataset of long-horizon manipulation trajectories featuring a Franka arm performing a kettle pouring task. Includes raw and processed trajectory variants with preprocessing and structure details. Serves as a reference dataset for imitation learning and physical AI data pipeline development.

  Source
• Axis Robotics releases AxisDataCleaning trajectory-processing toolkit

  2026-04-27 - Research Publication - The repository processes Axis teleoperation trajectories supplied as JSON or CSV, including multi-trajectory and batch inputs. Its pipeline combines static-frame merging, Savitzky-Golay smoothing and cubic-spline resampling, then reports smoothness and trajectory-deviation metrics. Offline task checkers and a benchmark script support repeatable validation of cleaned trajectories before training use.

  Source

Events

• 2026-09-04 - Research Publication - Axis Robotics releases Axis Sim Dataset V1

  The public release includes the dataset, training code and benchmarks for a simulated Franka Research 3 arm. Continual pretraining on V1 increased π0.5 success on LIBERO-Plus from 83.9% to 88.8%. The trajectories cover pick-and-place, stacking, pouring, articulated-object manipulation and tool use collected through Axis Hub.

  Source
• 2026-07-27 - Funding - Axis Robotics raises $12M seed for physical-AI data infrastructure

  Hack VC led the $12 million seed round, with Nomad Capital, Pi Network Ventures, 10K Ventures and angel investors participating. Axis said the financing would support its community-driven data engine for physical-AI training and continuous policy improvement. The company reported more than 100,000 contributors at announcement; that figure remains a company disclosure.

  Source
• 2026-07-23 - Research Publication - Axis Robotics publishes the AXIS community-driven robot-learning data engine

  The released AXIS dataset contains 207 manipulation tasks and more than 50,000 trajectories. The system combines browser-based teleoperation with automated task generation, success checking, trajectory cleaning and visual and physics-based augmentation. Continual pretraining improved the reported overall success rate of π0.5 by 5.8% and outperformed the RoboCasa365-pretrained comparison by 37.3%.

  Source
• 2026-07-10 - Product Launch - Axis Robotics releases Axis V2 for online post-training

  Axis V2 supports scalable online post-training for physical-AI models. The release belongs to the existing Axis V2 product rather than separate module products.

  Source
• 2026-04-29 - Research Publication - Axis Task 74 sample dataset – long-horizon manipulation (kettle pouring)

  Sample dataset of long-horizon manipulation trajectories featuring a Franka arm performing a kettle pouring task. Includes raw and processed trajectory variants with preprocessing and structure details. Serves as a reference dataset for imitation learning and physical AI data pipeline development.

  Source
• 2026-04-27 - Research Publication - Axis Robotics releases AxisDataCleaning trajectory-processing toolkit

  The repository processes Axis teleoperation trajectories supplied as JSON or CSV, including multi-trajectory and batch inputs. Its pipeline combines static-frame merging, Savitzky-Golay smoothing and cubic-spline resampling, then reports smoothness and trajectory-deviation metrics. Offline task checkers and a benchmark script support repeatable validation of cleaned trajectories before training use.

  Source
• 2026-03-12 - Product Launch - Axis Robotics opens its browser-based robot-training data platform

  The platform combines task generation, simulation, human demonstration collection and data processing. The official surface preserved beta and waitlist maturity rather than implying broad commercial deployment.

  Source

Relationships

• Geely Auto Group

  Manufacturing Sim To Real Collaborator - Axis collaborated with Geely Auto Group on simulated-to-real component sorting for automotive production.

• Booster Robotics

  Simulation Data Partner - Axis identifies Booster Robotics as its first simulation-data partner.