π0 (pi-zero)
No estimate
No hardware requirements for this model
The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.
On record
Full specification
Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.
Origin
Who built this model, where, and when it was published.
- Organisation
- Physical Intelligence
- Organisation type
- Industry
- Country
- United States of America
- Published
- 31 October 2024
- Authors
- Kevin Black, Noah Brown, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Lachy Groom, Karol Hausman, Brian Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Sergey Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Lucy Xiaoyang Shi, James Tanner, Quan Vuong, Anna Walling, Haohuan Wang, Ury Zhilinsky
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Robotics, Vision
- Task
- Robotic manipulation
- Base model
- PaliGemma
Size
How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.
- Parameters
- 3.3B
- Training data
- 1,000,000,000 tokens
"While in principle our model can be initialized from scratch or fine-tuned from any VLM backbone, in practice we use PaliGemma [5] as our base model. PaliGemma is an opensource 3 billion parameter VLM that offers a convenient tradeoff between size and performance. We add 300M parameters for the action expert (which is initialized from scratch) for a total of 3.3 billion parameters."
10,000 hours, and ~1 billion timesteps "We evaluate our approach by pre-training on over 10,000 hours of robot data, and fine-tuning to a variety of dexterous tasks" ... "We provide an overview of our pre-training mixture in Figure 4. Since each training example corresponds to a timestep — i.e., a tuple (ot, At), — we will quantify data in terms of timesteps in this discussion. 9.1% of the training mixture consists of open-source datasets, including OXE [10], Bridge v2 [52], and DROID [23]. Th…
Availability
Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.
- Weights
- Closed — provider access only
- Model access
- Unreleased
- Training code
- Unreleased
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
Sources
Where this record came from and when it was last checked.
- Reference
- π0: Our First Generalist Policy
- Last updated
- 11 February 2026
What the numbers mean
What this model is
π0 (pi-zero) was published by Physical Intelligence, in the country recorded as United States of America, during October 2024. It comes out of an organisation categorised as industry.
It works in the domain of Robotics, Vision, and is recorded as performing the task of robotic manipulation.
Its starting point was an existing base model, PaliGemma. That is why it shares the base model's general shape and size.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
How it was trained
The training set ran to roughly 1,000,000,000 tokens of text.
Answers
π0 (pi-zero) — common questions
π0 (pi-zero)— what GPU do I need to run it?
None. This is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.
π0 (pi-zero)— is it open source?
No. Its weights have not been published, so it exists only as a service controlled by its owner.
π0 (pi-zero)— how many parameters does it have?
It has a parameter count of 3.3B. "While in principle our model can be initialized from scratch or fine-tuned from any VLM backbone, in practice we use PaliGemma [5] as our base model. PaliGemma is an opensource 3 billion parameter VLM that offers a convenient tradeoff between size and performance. We add 300M parameters for the action expert (which is initialized from scratch) for a total of 3.3 billion parameters.". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.
π0 (pi-zero)— who created it?
It was published by Physical Intelligence, based in United States of America, an organisation categorised as industry.
π0 (pi-zero)— when was it released?
It was published in October 2024. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
π0 (pi-zero)— what is it used for?
It works in the domain of Robotics, Vision, and is recorded as handling the task of robotic manipulation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
The other direction
Looking at it from the other side?
This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.