Skip to content
embodied.guide

Back to the AI models

PerceptronUnited States2026

Isaac 0.5

An AI model: an embodied reasoning model reasons about what a robot sees and what to do next. Its answer is typically a plan, a location or text rather than motor commands. This page describes the software, not a robot.

Robotsnot stated

Key figures as stated by the developer

5 of 6 stated · 1 gap
  • Model type ⁠A embodied reasoning
  • Controls not stated
  • Inputs ⁠A camera language robot state
  • Access ⁠A open download
  • Parameters ⁠A 36,000,000,000
  • Base model ⁠A Qwen-family vision-language backbone (36B, sparse experts)
How to read the figures
50HzA
The letter in the grey square after a figure points to the source the figure came from. Pages with several sources carry several letters.

The four states

Versions and notes in the key figures (6)

Model type

Developer wording: "Isaac 0.5 brings multimodal video understanding, embodied reasoning, spatial grounding, task-progress estimation, and robot control into one 36-billion-parameter sparse model."

Controls

Note Model card: "It reads images, video, language instructions, robot state, and previous actions, then produces text, normalized coordinates, task-state outputs, or robot actions."

Inputs

Model card: "It reads images, video, language instructions, robot state, and previous actions, then produces text, normalized coordinates, task-state outputs, or robot actions." Images and video = camera, language instructions = language.

Access

Model card: "This model repository provides checkpoint weights and portable runtime manifests."

Parameters

Model card: "The checkpoint contains 36B parameters in total."

Base model

Model card, Model architecture: "Isaac uses a 36B Qwen-family vision-language backbone with sparse experts."

Full specifications

8 of 12 fields stated
The letter points to the source at the foot of the page. Fields without a figure are dashed: the developer does not state them — that is not the same as zero, and not the same as no.
Field Value Source and caveats
Input and output
Model typeembodied reasoningA

Developer wording: "Isaac 0.5 brings multimodal video understanding, embodied reasoning, spatial grounding, task-progress estimation, and robot control into one 36-billion-parameter sparse model."

Inputscamera language robot stateA

Model card: "It reads images, video, language instructions, robot state, and previous actions, then produces text, normalized coordinates, task-state outputs, or robot actions." Images and video = camera, language instructions = language.

Controlsnot stated

Note Model card: "It reads images, video, language instructions, robot state, and previous actions, then produces text, normalized coordinates, task-state outputs, or robot actions." ⁠A
Size and speed
Parameters36,000,000,000A

Model card: "The checkpoint contains 36B parameters in total."
Where it runs and what it is built on
Runs wherenot stated

Note Model card: the model can be "deployed as an action policy through LeRobot or our reference server". ⁠A

Base modelQwen-family vision-language backbone (36B, sparse experts)A

Model card, Model architecture: "Isaac uses a 36B Qwen-family vision-language backbone with sparse experts."
Openness
Accessopen downloadA

Model card: "This model repository provides checkpoint weights and portable runtime manifests."

LicenseApache-2.0A

Model card header: "license: apache-2.0".
The developer’s own claims
Learns new tasksneeds retrainingA

Developer claim: the model "Isaac 0.5 can be fine-tuned on new demonstrations".

Works across robotsyesB

Developer claim: "Isaac 0.5 supports cross embodiment"; the model card says "Isaac 0.5 is trained on more than 35 robot systems".

Yes only if the developer itself says the model runs on different robot models, for example robots from two manufacturers or two of one manufacturer’s models. Several units of the same robot do not count.

Not stated: Degrees of freedom controlled, Control rate

Notes

  • Perceptron describes Isaac 0.5 as its open foundation model for robot learning.
  • Perceptron released Isaac 0.5 on 26 August 2026 as an open-source embodied foundation model.

Sources

One line per source, not per figure: a source has one retrieval date. The letter beside a figure points to the line it came from.