PhysicalAI.bestOur standard
Imitation-learning policy

ACT

ACT predicts chunks of robot actions from demonstrations, with the original project focused on fine-grained bimanual manipulation.

Source contextPrimary specReviewed
Sources 2
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware

Tony Z. Zhao and coauthors · Publication date not disclosed · Source accessed 2026-09-28

Reviewed recordLast reviewed 2 sources ↗4 sourced fields
Official source Suggest a correction

Architecture and scope

The code supports simulated transfer-cube and insertion tasks and distinguishes real ALOHA execution from simulation. Camera setup and action representation belong to the trained policy configuration.

Source contextPrimary specReviewed
Sources 2
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware

Tony Z. Zhao and coauthors · Publication date not disclosed · Source accessed 2026-09-28

Integration and evaluation

Temporal ensembling is an explicit runtime option. It should be recorded alongside a checkpoint when comparing trajectories or latency rather than treated as an invariant property of all ACT deployments.

Source contextPrimary specReviewed
Sources 2
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware

Tony Z. Zhao and coauthors · Publication date not disclosed · Source accessed 2026-09-28

Evidence boundary

The public method and a user-trained task policy are different artifacts; no universal task success rate is assigned.

Source contextPrimary specReviewed
Sources 2
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware

Tony Z. Zhao and coauthors · Publication date not disclosed · Source accessed 2026-09-28

Model identity

Developer
Tony Z. Zhao and coauthors
Primary specReviewed
Sources 1
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Exact version / scope
Action Chunking with Transformers reference release
Primary specReviewed
Sources 1
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Category
Imitation-learning policy
Primary specReviewed
Sources 1
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Access model
Public code and research training workflows
Primary specReviewed
Sources 1
tonyzhaozh/act official repository

tonyzhaozh · Publication date not disclosed · Source accessed 2026-09-28

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

License
Not publicly disclosed

Inputs, outputs & embodiment

Inputs
Not publicly disclosed
Outputs
Not publicly disclosed
Embodiment
Not publicly disclosed

Implementation & training

Compute
Not publicly disclosed
Training-data disclosures
Not publicly disclosed
Integrations
Not publicly disclosed

Reported benchmark results

Context before scores. Results from different benchmarks are not directly comparable. A simulation result does not establish real-world reliability, safety or commercial availability.
No result meets our complete revision and methodology requirements for this record yet. Inspect the original benchmark documentation before comparing published scores.

What this evidence does not establish

  • The public method and a user-trained task policy are different artifacts; no universal task success rate is assigned.
  • Pricing and deployment service terms were not verified.

Relationships & deployments

stackHugging Face LeRobot ↗modelOpenVLA-OFT ↗modelDiffusion Policy ↗

Related reading is an editorial crosslink. Sourced connections describe relationships reported in the cited material. A link to a versioned profile does not establish compatibility with that version unless the connection note explicitly identifies it.

Record history & verified changes

A research review records when we checked a source. It does not mark a product launch or a new deployment.

Initial reviewed record. No subsequent field change has been recorded.

Inspect the evidence

Sources & evidence

md-act
tonyzhaozh/act official repository

tonyzhaozh · Repository · Publication date not disclosed · Source accessed 2026-09-28

Factual summary and attribution only; no upstream prose, images, weights, or dataset redistributed.

Official README inspected live. Repository code terms do not automatically cover model weights, data, or third-party assets.

Referenced by 1 record: ACT