Skip to content
RobothecaBeta
Log in

Isaac 0.5 (Perceptron)

  • AI model
  • Humanoids and robot learning

Isaac 0.5 is a 36-billion-parameter open model for video understanding, embodied reasoning, and robot control, released with training and inference code built on LeRobot.

Made by Perceptron AI, United States (checked on github.com/perceptron-ai-inc (opens in a new tab)).

Source on GitHub: github.com/perceptron-ai-inc/isaac (opens in a new tab). Facts read at commit be6507b (opens in a new tab).

Licenses

Code
Apache-2.0 (opens in a new tab)

Read by Robotheca from LICENSE (opens in a new tab) (Apache-2.0) at commit be6507b (opens in a new tab), on 16 Oct 2026.

Commercial use
Yes

What it does

  • Reads images, video, language instructions, robot state and past actions, and answers questions about video.
  • Points to and tracks objects, estimates task progress, and generates robot actions.
  • Trained on data from more than 35 robot systems, 100,000 hours of robot experience, and one million hours of general video.
  • Offers continuous control via a Flow expert and diffusion transformer, and discrete control via 2,048 FAST tokens.
  • Predicts the next action chunk during closed-loop control while the current chunk is still running.
  • Ships a policy type named perceptron_isaac in the LeRobot integration, with fine-tuning and runtime guides.

Written by Robotheca from the project's README (opens in a new tab) at commit be6507b.

Getting started

Difficulty
Advanced

Maintenance

Status
Active: a commit in the last 90 days
Newest commit
26 Aug 2026 on main
Latest release
None published on GitHub

Checked on GitHub on 16 Oct 2026.

Links

Robots that can run it

No robot on Robotheca is linked to it yet. Links come from a robot's robot.yaml, from its dependency files read as text, or from Robotheca's editors.

Robotheca lists AI models as a reference: nothing is hosted or run here, and every fact links to where it was read.