GAIA-2
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
- Wayve
- Organisation type
- Industry
- Country
- United Kingdom of Great Britain and Northern Ireland
- Published
- 26 March 2025
- Authors
- Lloyd Russell, Anthony Hu, Lorenzo Bertoni, George Fedoseev, Jamie Shotton, Elahe Arani, Gianluca Corrado
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Video, Vision, Multimodal, Language
- Task
- Self-driving car, Video generation, Instruction interpretation
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
- 8.7B
- Training data
- tokens
video tokenizer: 285M parameters workd model: 8.4B parameters total: 8.685B parameters
"The dataset comprises approximately 25 million video sequences, each spanning 2 seconds, collected between 2019 and 2024. Recordings were obtained across three countries—the United Kingdom, the United States, and Germany—to ensure coverage of geographically and environmentally diverse driving conditions. "
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
- 7.6 × 10²² FLOP
- How it was established
- Operation counting
"The video tokenizer was trained for 300,000 steps with a batch size of 128 using 128 H100 GPUs. Input sequences consisted of Tv = 24 video frames sampled at their native capture frequencies (20, 25, or 30 Hz). Random spatial crops of size 448 × 960 were extracted from the frames. For each training sample, a camera view was randomly selected from the available N = 5 perspectives." "The latent world model was trained for 460,000 steps with a batch size of 256 on 256 H100 GPUs. Inputs consisted o…
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
- NVIDIA H100 SXM5 80GB
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
- GAIA-2: A Controllable Multi-View Generative World Model for Autonomous Driving
- Last updated
- 28 November 2025
What the numbers mean
Where it came from
GAIA-2 was published by Wayve, in the country recorded as United Kingdom of Great Britain and Northern Ireland, during March 2025. It comes out of an organisation categorised as industry.
It works in the domain of Video, Vision, Multimodal, Language, and is recorded as performing the task of self-driving car, Video generation, Instruction interpretation.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
What went into building it
The training run consumed about 7.6 × 10²² FLOP, on hardware recorded as NVIDIA H100 SXM5 80GB. That figure measures what producing the model cost, and has no bearing on how fast it answers.
Answers
GAIA-2 — common questions
GAIA-2— how many parameters does it have?
It has a parameter count of 8.7B. video tokenizer: 285M parameters workd model: 8.4B parameters total: 8.685B 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.
GAIA-2— who created it?
It was published by Wayve, based in United Kingdom of Great Britain and Northern Ireland, an organisation categorised as industry.
GAIA-2— when was it released?
It was published in March 2025.
GAIA-2— what is it used for?
It works in the domain of Video, Vision, Multimodal, Language, and is recorded as handling the task of self-driving car, Video generation, Instruction interpretation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.
GAIA-2— how much compute was used to train it?
Training consumed around 7.6 × 10²² FLOP, on hardware recorded as NVIDIA H100 SXM5 80GB. 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.
GAIA-2— 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.
GAIA-2— is it open source?
No. Its weights have not been published, so it exists only as a service controlled by its owner.
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.