FTW (For The Win)
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
- 3 July 2018
- Authors
- Max Jaderberg, Wojciech M. Czarnecki, Iain Dunning, Luke Marris, Guy Lever, Antonio Garcia Castaneda, Charles Beattie, Neil C. Rabinowitz, Ari S. Morcos, Avraham Ruderman, Nicolas Sonnerat, Tim Green, Louise Deason, Joel Z. Leibo, David Silver, Demis Hassabis, Koray Kavukcuoglu, Thore Graepel
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Games
- Task
- Capture the flag
- Approach
- Self-supervised learning
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
- 126M
- Training data
- 2,000,000,000 tokens
Architecture described in figure S11 of the supplement The architecture includes modules for visual embedding, reward prediction, recurrent processing, policy, baseline and pixel control. Input is 84x84x3 pixels as seen in figure S10 of the supplement "We elected to use a resolution of 84x84 pixels as in previous related work in this environment. Each pixel is represented by a triple of three bytes" Visual embedding (84x84x3 -> 256) 32*(8*8*3+1)+64*(4*4*32+1)+64*(3*3*64+1)+64*(3*3*64+1) + (8…
Agents were trained for two billion steps, corresponding to approximately 450K games.
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
- 3.5 × 10¹⁹ FLOP
- How it was established
- Third-party estimation
Source: https://docs.google.com/spreadsheets/d/1Kj4Q5WADcDXtUJLIOfGTCE3tGvxNczEMwyy8QtgSkHk/edit#gid=54587040&fvid=1361937389 Additional estimate based on parameters (low confidence, had to make some assumptions about the architecture) Forward pass operations Vision 84*84*32*8*8*3+20*20*32*64*4*4+7*7*64*64*3*3+7*7*64*64*3*3+7*7*64*256=60874752 60874752*2=121749504 Pixel control (guess on the internal dimensions) 256*32*7*7+7*7*32*32*9*9+20*20*32*4*4*(1+2+3+4+5)=7537664 Remaining 2*(463616+331…
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.
- Why it is tracked
- SOTA improvement
- Record confidence
- Speculative
- Citations
- 802
"In this work, we demonstrate for the first time that an agent can achieve human-level in a popular 3D multiplayer first-person video game, Quake III Arena Capture the Flag (28), using only pixels and game points as input."
Sources
Where this record came from and when it was last checked.
- Reference
- Human-level performance in first-person multiplayer games with population-based deep reinforcement learning
- Last updated
- 25 May 2026
What the numbers mean
Where it came from
FTW (For The Win) was published by DeepMind, in United Kingdom of Great Britain and Northern Ireland, in July 2018. The organisation is categorised as industry.
It works in Games, and is recorded as doing capture the flag.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
Training and provenance
Training it took roughly 3.5 × 10¹⁹ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.
The training set ran to roughly 2,000,000,000 tokens.
The reason it appears in this catalogue at all is sOTA improvement.
Answers
FTW (For The Win) — common questions
How much compute was used to train FTW (For The Win)?
Around 3.5 × 10¹⁹ 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 FTW (For The Win)?
None. FTW (For The Win) 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 FTW (For The Win) open source?
No. FTW (For The Win) has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does FTW (For The Win) have?
FTW (For The Win) has 126M parameters. Architecture described in figure S11 of the supplement The architecture includes modules for visual embedding, reward prediction, recurrent processing, policy, baseline and pixel control. Input is 84x84x3 pixels as seen in figure S10 of the supplement "We elected to use a resolution of 84x84 pixels as in previous related work in this environment. Each pixel is represented by a triple of three bytes" Visual embedding (84x84x3 -> 256) 32*(8*8*3+1)+64*(4*4*32+1)+64*(3*3*64+1)+64*(3*3*64+1) + (84/(S^4)*84/(S^4)*64+1)*256 Note there is no information about the stride S used in the convolutions; we assume S = 1 Reward prediction (256 -> 3) (256+1)*128 + (128+1)*3 Recurrent processing (n-> 512) VU1 (256 -> 512) 4*(799+2*32)*((512+(32*2) + 3*32 + 5*2 + 3)+(799+2*32)+1) + 2*(256+1)*256 VU2 (512 -> 512) 4*(512+2*32)*((512+(32*2) + 3*32 + 5*2 + 3)+(512+2*32)+1) + 2*(256+1)*256 LSTMs usually have 4*(n*m+n*n+n) parameters, where n=input size and m=output size. This DNS + LSTM takes as input the concatenation of the previous layer of size n and R read vectors of size W=32; and outputs m units plus an interface vector of size (W*R) + 3*W + 5*R + 3, for a total of about 4*(n+R*W)*((m+(W*R) + 3*W + 5*R + 3)+(n+R*32)+1) parameters I assume R=2 since that seems implied by the previous paper (?) The first VU has as input the visual embedding (size 256), the previous action (size 540) and the previous reward (size 3), for a total size of 256+540+3 = 799. The output is size 512. The second VU has input size 512 and output size 512 The DNC memory architecture is described in https://www.nature.com/articles/nature20101.epdf Policy (512 -> 5x3x3x3x2x2) 6*(512+1)*256 + (256+1)*5 + 3*(256+1)*3 + 2*(256+1)*2 Baseline (512+1)*256 + (256+1)*1 Pixel control (512+1)*32*7*7 + 32*(9*9+1) + 5*(4*4+1) + 3*2*(4*4+1) + 2*2*(4*4+1) + 1*(4*4+1) "we trained independent pixel control policies for each of the six action groups". 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 FTW (For The Win)?
FTW (For The Win) was published by DeepMind, based in United Kingdom of Great Britain and Northern Ireland, categorised as industry.
When was FTW (For The Win) released?
FTW (For The Win) was published in July 2018. 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 FTW (For The Win) used for?
FTW (For The Win) works in Games, and is recorded as handling capture the flag. These are the areas it was designed around; they describe intent rather than a hard boundary.
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.