Hide and Seek
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
- OpenAI
- Organisation type
- Industry
- Country
- United States of America
- Published
- 17 September 2019
- Authors
- Bowen Baker, Ingmar Kanitscheider, Todor Markov, Yi Wu, Glenn Powell, Bob McGrew, Igor Mordatch
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Games
- Task
- Hide and seek
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
- 1.6M
- Training data
- 120,000,000,000 tokens
"The default model, which uses a batch size of 64,000 and 1.6 million parameters,..." pg. 7 of the Hide and Seek paper, stored in "RL papers" source: https://docs.google.com/spreadsheets/d/1Kj4Q5WADcDXtUJLIOfGTCE3tGvxNczEMwyy8QtgSkHk/edit#gid=54587040&fvid=1361937389
"The default model, which uses a batch size of 64,000 and 1.6 million parameters, requires 132.3 million episodes (31.7 billion frames) over 34 hours of training to reach stage 4 of the skill progression, i.e. ramp defense." But the full number of episodes, e.g. Figure 1, is 500 million. 500 / 132.3 * 31.7B ~= 120B.
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
- 1.2 × 10¹⁸ FLOP
- How it was established
- Operation counting,Third-party estimation
1.6e6 * 2 * 120e9 * 3 ~= 1.15e18 FLOP. We assume the single convolution on the lidar input is negligible, and the rest of the model (which consists of MLPs, self-attention and LSTM) has roughly 1 multiply-add per parameter. The following source has a similar estimate but does not adjust for the full number of episodes: https://docs.google.com/spreadsheets/d/1Kj4Q5WADcDXtUJLIOfGTCE3tGvxNczEMwyy8QtgSkHk/edit#gid=54587040&fvid=1361937389
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
code for environment, but don't see training/inference code, MIT: https://github.com/openai/multi-agent-emergence-environments
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
- Citations
- 752
Sources
Where this record came from and when it was last checked.
- Reference
- Emergent Tool Use From Multi-Agent Autocurricula
- Last updated
- 25 May 2026
What the numbers mean
Background
Hide and Seek was published by OpenAI, in United States of America, in September 2019. industry is the category the publisher falls under.
It works in Games, and is recorded as doing hide and seek.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
Training and provenance
Producing it required around 1.2 × 10¹⁸ FLOP of arithmetic, which is a statement about the training budget rather than about inference.
It was trained on about 120,000,000,000 tokens of text.
Answers
Hide and Seek — common questions
What is Hide and Seek used for?
Hide and Seek works in Games, and is recorded as handling hide and seek. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.
How much compute was used to train Hide and Seek?
Around 1.2 × 10¹⁸ FLOP. 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 Hide and Seek?
None. Hide and Seek 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 Hide and Seek open source?
No. Hide and Seek has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does Hide and Seek have?
Hide and Seek has 1.6M parameters. "The default model, which uses a batch size of 64,000 and 1.6 million parameters,..." pg. 7 of the Hide and Seek paper, stored in "RL papers" source: https://docs.google.com/spreadsheets/d/1Kj4Q5WADcDXtUJLIOfGTCE3tGvxNczEMwyy8QtgSkHk/edit#gid=54587040&fvid=1361937389. 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 Hide and Seek?
Hide and Seek was published by OpenAI, based in United States of America, categorised as industry.
When was Hide and Seek released?
Hide and Seek was published in September 2019. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
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.