AI models
GameNGen
Published
Aug 22, 2024
GameNGen is an AI model developed by Google Research, Google DeepMind and Tel Aviv University (United States and Israel), first published in August 2024. It works in the games domain, on tasks such as doom.
Epoch AI has no training-compute estimate for this model. It was trained on roughly 458.8B datapoints. Training ran on 128 Google TPU v5e.
Access: Unreleased. Its weights are not openly released. It is built on top of Stable Diffusion (LDM-KL-8-G).
Full record
- Organization
- Google Research, Google DeepMind, Tel Aviv University
- Country of organization
- United States, Israel
- Domain
- Games
- Task
- Doom
- Dataset size
- 458.8B
- Training hardware
- Google TPU v5e
- Chips used
- 128
- Training power draw
- 56.8 kW
- Model accessibility
- Unreleased
- Open weights
- No
- Base model
- Stable Diffusion (LDM-KL-8-G)
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models