Octo
Robotics modelTransformer diffusion robot policies with pretrained checkpoints and fine-tuning tools; observation history, action statistics and robot-specific interfaces must match the selected model. [1]
Setup & usage →CheckpointsModel hub / homepageGetting started
- For
- Robot-learning researchers with a compatible environment and accelerator setup [1]
- Embodiments
- WidowX (BridgeData V2), ALOHA (via fine-tuning example), Open X-Embodiment 25-dataset mixture robots (Franka, UR5,… [7] [8]
- Source rights
- Some licenses are unresolved. Review licenses
Source researched · Not tested by us
Sources checked · Research details
Licenses & related hardware
- License
- MIT · license not fully assessed for datasets, documentation · Details by part
Overview
Octo (octo-models/octo, release v1.5) is a family of generalist transformer diffusion visuomotor policies trained on a 25-dataset mixture from Open X-Embodiment. Two ungated JAX Orbax checkpoints are published on Hugging Face Hub: rail-berkeley/octo-base-1.5 (93M parameters, ViT-B equivalent, 0.75 GB on disk) and rail-berkeley/octo-small-1.5 (27M parameters, ViT-S equivalent, 0.51 GB on disk). Both tokenize images via a lightweight CNN into 16x16 patches, encode language with T5-Base, and predict continuous actions with a diffusion head. [1] [2] [8] [9] [10]
Model card summary
- Architecture
Transformer-based generalist diffusion policy (ViT-B 93M for Base, ViT-S 27M for Small) using a lightweight CNN image tokenizer (16x16 patches), T5-Base language encoder, and diffusion action head.
- Base model
T5-Base language encoder (vision encoder and transformer trained from scratch on OXE)
- Input modalities
- rgblanguage
Primary RGB (256x256), wrist RGB (128x128), goal images, and language instructions (history window up to 2 timesteps); fine-tuning can add proprioception.
- Action space
Continuous diffusion action chunks predicting 4 future steps of 7-dimensional actions (window size 2; translation, rotation, and gripper in the Bridge example), unnormalized via dataset/robot-specific statistics; supports full-chunk, first-step, or temporal-ensembling execution.
- Trained / supported embodiments
- WidowX (BridgeData V2)
- ALOHA (via fine-tuning example)
- Open X-Embodiment 25-dataset mixture robots (Franka, UR5, Kuka, Google Robot, Stretch, Jaco, etc.)
- Training datasets
- Open X-Embodiment 25-dataset mixture (Fractal, Kuka, Bridge, BC-Z, Stanford Hydra, Language Table, etc.)
Checkpoints
| Checkpoint | Format | Size | Access | Pinned revision |
|---|---|---|---|---|
| rail-berkeley/octo-base-1.5 | jax-orbax | 0.75 GB | open | ee3c10e8edd6 |
| rail-berkeley/octo-small-1.5 | jax-orbax | 0.51 GB | open | dc9aa3019f76 |
Ungated JAX Orbax checkpoints verified via Hugging Face API.
Is this for you?
- Use cases
- Zero-shot language- or goal-image-conditioned manipulation on OXE-matched setups
- Fine-tuning compact diffusion policies for new observation/action spaces (e.g., ALOHA)
- Known limitations
- Action chunks are not complete task trajectories
- Missing modalities and absent history require appropriate masks
- Pretraining dataset and compute are substantially larger than the inference example
- Model and source licenses do not establish a uniform license for every training dataset
Last item is an editorial rights-scope limitation.
- Safety and limitations
- 4-step action chunks are not complete task trajectories; execution requires matching history masks (timestep_pad_mask) and per-modality masks.
- Pinned to Python 3.10 and JAX 0.4.20 (CUDA 11 or TPU), which requires care on newer CUDA/driver stacks.
- MIT code and checkpoint licenses do not establish a uniform license across the 25 constituent Open X-Embodiment datasets.
Compute & fine-tuning
- Inference compute
- Documented hardware: CUDA 11 GPU or TPU (JAX 0.4.20; 0.51-0.75 GB checkpoints)
- Precision:
fp32 - Edge / consumer capable: Yes
Compact 27M-93M parameter footprint (0.51-0.75 GB on disk); edge/single-GPU inference capability is editorially inferred while exact minimum VRAM and latency_ms are not stated.
- Fine-tuning
- Supported: Yes
- Methods: Full or observation/action-adapter fine-tuning in JAX
- Recipe: Upstream fine-tuning documentation
Debug fine-tuning example loads octo-small-1.5; ALOHA example adds separate simulator and ACT dependencies.
- Host computer
CUDA 11 GPU or TPU with JAX 0.4.20 for inference and fine-tuning; upstream pretraining used ~1.2 TB prepared data on TPU pods.
- Operating system
- Linux
CUDA 11 / TPU JAX Linux setup documented.
[1] · Editorial assessment
- Dependencies
- Python 3.10 in upstream environment instructions
- JAX 0.4.20 with CUDA11 or TPU installation route
- Checkpoint and robot/dataset-specific normalization statistics
- Separate robot or Gym-compatible evaluation environment
Setup & usage
Open the upstream getting-started guide [5] · Model card / documentation [1] [4]
- Create Python 3.10 environment and install JAX 0.4.20 (CUDA 11 or TPU route)
- Load a pinned v1.5 checkpoint (octo-base-1.5 or octo-small-1.5) via OctoModel.load_pretrained
- Validate observation keys, history masks (timestep_pad_mask), and dataset unnormalization statistics using 01_inference_pretrained.ipynb
Upstream-reported benchmarks
Upstream-reported only: Open Source Robotics has not run training, inference, or hardware evaluations for these benchmarks.
| Benchmark | Upstream-reported score | Scope & caveats |
|---|---|---|
| Octo upstream paper and repository evaluations | Upstream reports multi-embodiment zero-shot and fine-tuning results in the project paper and README (specific scores not reproduced here) | Upstream-reported evaluation claims; not independently verified by OSR. |
Project notes
Observation history, masks, and action chunks
Both v1.5 checkpoints accept primary RGB images (256x256), wrist RGB images (128x128), goal images, and language instructions over a history window of up to 2 timesteps, predicting 4-step chunks of 7-dimensional actions. In the supplied Bridge inference notebook (01_inference_pretrained.ipynb), those seven components map to end-effector translation, rotation, and gripper control; target environments must match the selected checkpoint's action convention and dataset unnormalization statistics rather than assuming identical semantics across robots. At inference, any subset of observation and task keys may be passed, but timestep_pad_mask must mark absent history steps separately from missing camera modalities, and chunks can be executed via full-chunk, first-action, or temporal-ensembling loops. Consult the Gym adapter guide (examples/envs/README.md) as the next integration reference when wiring a new environment. [1] [5] [6] [8]
Environment, fine-tuning, and upstream-reported benchmarks
Upstream setup specifies Python 3.10 and JAX 0.4.20 with CUDA 11 or TPU routes. Upstream documents GPU/TPU inference as well as fine-tuning workflows for new observation and action spaces (including an ALOHA example in 03_eval_finetuned.py with separate simulator and ACT dependencies), while full pretraining from scratch uses ~1.2 TB of prepared data on TPU pods. Minimum VRAM, inference latency, and target-robot execution timing/control frequency are not established here—checkpoint download sizes (0.51–0.75 GB) are disk sizes rather than runtime memory requirements, and single-GPU suitability for fine-tuning is not established by the inspected sources. Upstream evaluation claims in the project paper and README are upstream-reported; Open Source Robotics has not downloaded checkpoints or run training, fine-tuning, inference, or robot evaluations. [1] [7] [10]
What we checked
Source research only; no installation, physical build or performance test performed. No affiliate links added.
Not yet established
- Documented closed-loop control frequency (Hz)
Updates
Separate license matrix
Available project sources support research coverage; complete build and commercial qualification remain separate.
| Layer | Availability / license |
|---|---|
| Code | available · MIT [3] |
| Model weights | available · MIT Pinned Octo Base 1.5 (ee3c10e8) and Octo Small 1.5 (dc9aa301) model cards both declare MIT; checkpoint payloads not downloaded or executed. [1] [8] [9] [10] |
| Training datasets | available · License not fully assessed Upstream links the Open X-Embodiment collection and preparation script. Constituent dataset licenses and payloads were not individually verified; no uniform dataset license asserted. [1] [8] |
| Documentation | available · License not fully assessed Examples and source documentation available; separate documentation-wide license not established. [1] [4] [5] |
Separate code, weight, and dataset rights
The octo-models/octo repository and both pinned v1.5 model cards (octo-base-1.5 and octo-small-1.5) declare MIT licenses. However, the 25 constituent Open X-Embodiment datasets retain their own upstream licenses, which were not individually verified here. [1] [3] [8] [9]
References
Each source links to the exact revision we read. “Used for” shows where this page relies on it.
- S001 octo-models/octo: README.md
Revision, date & where cited
github.com · revision
5eaa5c6· Accessed Oct 1, 2026
Used for: Overview, Model card summary, Is this for you?, Compute & fine-tuning and 5 more - S002 octo-models/octo: commit 5eaa5c6
- S003 octo-models/octo: LICENSE
Revision, date & where cited
github.com · revision
5eaa5c6· Accessed Oct 1, 2026
Used for: Is this for you?, Separate license matrix - S004 octo-models/octo: files
Revision, date & where cited
github.com · revision
5eaa5c6· archive file inventory · Accessed Oct 1, 2026
Used for: Setup & usage, Separate license matrix - S005 Octo v1.5: examples/01_inference_pretrained.ipynb
Revision, date & where cited
github.com · revision
5eaa5c6· Accessed Oct 1, 2026
Used for: Model card summary, Is this for you?, Compute & fine-tuning, Setup & usage and 2 more - S006 Octo v1.5: examples/envs/README.md
Revision, date & where cited
github.com · revision
5eaa5c6· Accessed Oct 1, 2026
Used for: Model card summary, Compute & fine-tuning, Observation history, masks, and action chunks - S007 Octo v1.5: examples/03_eval_finetuned.py
Revision, date & where cited
github.com · revision
5eaa5c6· Accessed Oct 1, 2026
Used for: Model card summary, Is this for you?, Compute & fine-tuning, Environment, fine-tuning, and upstream-reported benchmarks - S008 Octo Base 1.5 checkpoint model card
Revision, date & where cited
octo-models / project contributors · huggingface.co · README.md · Accessed Oct 1, 2026
Used for: Overview, Model card summary, Checkpoints, Is this for you? and 2 more - S009 Octo Small 1.5 checkpoint model card
27M-parameter ViT-S equivalent Transformer diffusion policy with CNN image tokenizer and T5-Base language encoder, predicting 4-step chunks of 7-D actions, license: mit.Revision, date & where cited
UC Berkeley RAIL (Hugging Face) · huggingface.co · README.md · Accessed Nov 21, 2026
Used for: Overview, Model card summary, Checkpoints, Separate license matrix - S010 Hugging Face API metadata: rail-berkeley/octo-base-1.5 and octo-small-1.5
Ungated Orbax checkpoints (300000/default/checkpoint + config.json + dataset_statistics.json): octo-base-1.5 is 0.75 GB and octo-small-1.5 is 0.51 GB, both license: mit, lastModified 2024-05-21.Revision, date & where cited
Hugging Face (rail-berkeley) · huggingface.co · HF API metadata for octo-base-1.5 (ee3c10e8) and octo-small-1.5 (dc9aa301) · Accessed Nov 21, 2026
Used for: Overview, Checkpoints, Compute & fine-tuning, Environment, fine-tuning, and upstream-reported benchmarks and 1 more
Correction guidance · Discussion is not enabled · Profile format 0.8.0-draft.1
Change history
2026-11-21 — Migrated 0.6.0-draft.1 -> 0.7.0-draft.5 -> 0.8.0-draft.1 (kind=model); added verified Small 1.5 card (S009) and HF API metadata (S010) while preserving historical source dates.