Gato
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
- DeepMind
- Organisation type
- Industry
- Country
- United Kingdom of Great Britain and Northern Ireland
- Published
- 12 May 2022
- Authors
- Scott Reed, Konrad Żołna, Emilio Parisotto, Sergio Gómez Colmenarejo, Alexander Novikov, Gabriel Barth-Maron, Mai Giménez, Yury Sulsky, Jackie Kay, Jost Tobias Springenberg, Tom Eccles, Jake Bruce, Ali Razavi, Ashley Edwards, Nicolas Heess, Yutian Chen, Raia Hadsell, Oriol Vinyals, Mahyar Bordbar, Nando de Freitas
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Multimodal, Robotics, Games, Language
- Task
- Atari, Image captioning, Chat, Robotic manipulation
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
- 1.2B
- Training data
- 524,288,000,000 tokens
"This section focuses on in-simulation evaluation. Figure 10 compares the full 1.18B parameter Gato" p.10
Training compute
The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.
- Training compute
- 4 × 10²¹ FLOP
- How it was established
- Hardware,Operation counting
256 (16x16x) TPUv3 chips x 123e12 FLOPS/chip x 4 days x 86400 seconds/day * 0.4 utilization = 4.35e21 FLOPs Similar value by 6NC: 6 * 524288000000 * 1.18B = 3.71e21 Using geometric mean: sqrt(4.35e21 * 3.71e21) = 4.02e21
The training run
What it physically took to train: which chips, how many, for how long, and what that drew from the wall.
- Training hardware
- Google TPU v3
- Chips used
- 256
- Wall-clock time
- 96 hours
- Power draw
- 231.3 kW
- Compute cost
- $3,523
4 days
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.
- Foundation model
- Yes
- Why it is tracked
- SOTA improvement
- Record confidence
- Confident
- Citations
- 1,064
SOTA at Meta-World MT50 tasks (96.6%) page 14, section 5.5
Sources
Where this record came from and when it was last checked.
- Reference
- A Generalist Agent
- Last updated
- 25 May 2026
What the numbers mean
Background
Gato was published by DeepMind, in United Kingdom of Great Britain and Northern Ireland, in May 2022. The organisation is categorised as industry.
It works in Multimodal, Robotics, Games, Language, and is recorded as doing atari, Image captioning, Chat, Robotic manipulation.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
How it was trained
Producing it required around 4 × 10²¹ FLOP of arithmetic, on Google TPU v3, which is a statement about the training budget rather than about inference.
The training set ran to roughly 524,288,000,000 tokens.
Its inclusion criterion is sOTA improvement.
Answers
Gato — common questions
How much compute was used to train Gato?
Around 4 × 10²¹ FLOP, on Google TPU v3. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.
What GPU do I need to run Gato?
None. Gato 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.
Is Gato open source?
No. Gato has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does Gato have?
Gato has 1.2B parameters. "This section focuses on in-simulation evaluation. Figure 10 compares the full 1.18B parameter Gato" p.10. 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.
Who created Gato?
Gato was published by DeepMind, based in United Kingdom of Great Britain and Northern Ireland, categorised as industry.
When was Gato released?
Gato was published in May 2022. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
What is Gato used for?
Gato works in Multimodal, Robotics, Games, Language, and is recorded as handling atari, Image captioning, Chat, 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.